CA2496155A1

CA2496155A1 - Compositions and methods for inferring ancestry

Info

Publication number: CA2496155A1
Application number: CA 2496155
Authority: CA
Inventors: Tony N. Frudakis; Mark D. Shriver
Original assignee: Individual
Current assignee: DNAPrint Genomics Inc
Priority date: 2002-08-19
Filing date: 2003-08-19
Publication date: 2004-02-26
Also published as: EP1578944A4; EP1578944A2; JP2012050432A; AU2003265572A1; WO2004016768A3; JP2006514553A; WO2004016768A2; AU2009225275A1

Abstract

Ancestry informative marker (AIMs) polynucleotides that contain a single nucleotide polymorphism or that contain an insertion or deletion, are provided, as are methods of using panels of the AIMs to draw an inference as to a trait of an individual. The trait can be, for example, biogeographical ancestry, a pigmentation trait, responsiveness to a drug, or susceptibility to a disease. Also provided are methods of determining the proportional ancestry of an individual. Reagents and kits also are provided.

Description

COMPOSITIONS AND METHODS FOR INFERRING ANCESTRY
BACKGROUND OF THE INVENTION
FIELD OF THE INVENTION
[0001] The invention relates generally to the identification of genetic markers predictive of an individual's biogeographical ancestryl and more specifically to combinations of single nucleotide polymorphisms useful as ancestry informative markers (AIMS), which allow an inference as to a trait of an individual, algorithms fox identifying such AIMS, and methods of using such AIMs to infer a trait of an individual, including an individual's ancestry, responsiveness of an individual to a drug, arid predisposition of an individual to a disease.
BACKGROUND INFORMATION

[0002] The majority (80-90%) of the genetic variation among human individuals is inter-individual, and only a relatively small proportion (10-20%) is due to population differences (Nei, In Molecular Population Genetics (Columbia University Press, New Yorlc) 1987;
CavaIIi-Sforza et aL, In The Histofy and Geography of Human Gees (Princeton University Press, Princeton NJ) 1994; Deka et al., Electr~opho~°esis 16:1659-1664, 1995; Rosenberg et al., Science 298:2381-2385, 2002; Akey et al., BioTechaiques 30:348-367, 2001;
Akey et al., Hum. Genet. 108:516-520, 2002). Most populations share alleles and those alleles that are most frequent in one population are also frequent in others. There are very few classical markers (e.g., blood group, serum protein, and immunological markers) or DNA
genetic marlcers that are population-specific or have large frequency differentials among geographically and ethnically defined populations (Roychodhury and Nei, In Humay2 Polymorphic Genes: World Dist~°ibution (Oxford University Press, New York) 1988; Dean et al., AnZer. J. Hum. Genet. 55:788-808, 1994; Cavalli-Sforza et al., supra, 1994, Akey et al., supra, 2001, 2002). Despite this apparent lack of unique genetic markers, there are marked physical and physiological differences among human populations that presumably reflect genetic adaptation to unique ecological conditions, random genetic drift, and sex selection.
In contemporary populations, these differences are evident in morphological differences between ethnic groups, as well as in differences in drug responsiveness and in susceptibility and resistance to disease.

[0003] On a basic level, human population structure can be represented in terms of BioGeographical Ancestry (BGA), which is the heritable component of "race" or heritage, and which is relevant on any scale of resolution. For example, on a crude level, BGA can be determined for 2 groups (e.g., European vs. others); or on a fine level, e.g., it can refer to "race" in terms of 4 groups such as IndoEuropeans, East Asians, sub-Saharan African and Native American; or on a finer level, e.g., it can refer to ethnicity within the European group (for example, Mediterranean or Scandinavian); or on a still finer level, e.g., it can even refer to groups of families within ethnic groups, such as groups of O'Reilly's descendent from a set of common ancestors within the Irish group. The measurement of BGA is relevant for most any type of genetics or epidemiological study design. For example, BGA is an important component in the variability of drug response (Burroughs et al., J. Natl. Med.
Assoc. 94:1-26, 2002). The reason for this relationship is that genetic dxift, geographical and/or reproductive isolation, and regional selective pressures have molded the allele frequencies of our ancestors for compatibility with allcaloids, tannins (self defense chemicals), and other xenobiotics fond in indigenous diets. Most drugs are derived from such chemicals and, therefore, it is no coincidence that the family of enzymes that allow humans to detoxify drugs are found at different frequencies in different populations. This scenario is not unique to drug responsiveness, and many other parts of the genome that are unrelated to drug responsiveness are subject to these same types of pressures.

[0004] Investigators generally have been concerned with identifying gene variants that cause a disease (the so called "phenotypically active" loci), rather than identifying gene variants that are simply correlated with disease. As such, whatever the trait being examined, and for most study designs involving unrelated individuals, it has been considered important to control for population structure so as to avoid identifying markers of structure that correlate with trait value in a given sample rather than those in linkage disequilibrium (LD) with phenotypically active loci (Risch et al., Geno~ze Biology 3:1-12, 2001 Wang et al., Anae~. J. Huf~a. Genet. 71:1227-1234, 2002; Burroughs et al., supra, 2002; Rao and Chakraborty, Amen. J. Hum. Genet. 26:444-453, 1974). There are two sources of population structure in a sample collection: 1) sampling effects, which can create structure even if sampling is performed from homogeneous populations, and 2) natural human demography.

The first source of population structure is a nuisance for genetics studies, and associations found from a study due to this type of structure are generally considered an artifact of the collection process rather than a reflection of human demography. Most geneticists generally consider the second lcind of structure to be a nuisance as well. As such, associations identified as being due to population structure have been considered spurious findings or artifacts, and have generally been discarded; only findings due to true linkage or LD have been published, as such marlcers are considered linked to biologically relevant genes.

[0005] Much effort has been directed to quantifying both types of population structure (above) in groups of individuals. Such methods essentially measure the departure from expected levels of heterozygosity within a group of samples as an indication of structure (though none of these methods are capable of xeading within-individual structure). Many common diseases exhibit locus and/or allelic heterogeneity as a function of BGA, and many authors have suggested that inappropriate attention to population structure during the study design step has produced at least some of the so-called "false positive"
results implicated in the rash of irreproducible Common Disease/Common Variant results obtained to date (Terwilliger et al., Curr. Opifz. Genet. bevel. 12:726-734, 2002). In order to control for the influence of population structure, several tests are appropriate (Cockerham, Evolution 23:72-83, 1969; Coclcerham, Genetics 74:679-700, 1973; Wier and Cockerham, Evolution 38:1358-1370, 1984; Long, Genetics 112:629-647, 1986, Excoffier et al., Genetics 131:343-359, 1992). These methods can be grouped in two main categories -genomic control methods (Devlin and Roeder, Biometrics 55:997-1004, 1999), and structured association (SA) methods (Pritchard and Donelly, Theo~. Popul. Biol. 60:227-237, 2001).
Both methods require genotyping of a panel of unlinlced markers to estimate and correct for the effect of genetic structure, but they axe usually applied for sample collections. However, should a pool of samples fail such a test, it is usually not clear which samples should be eliminated to rectify the problem. An equally vexing problem with this method is that it is often applied for a study sample after the creation of expensive data, thus creating a circular logic problem in addition to an economic problem; these methods axe usually employed to extract information on population structure using the characteristics of the data within which associations are sought.

[0006] In order to minimize the influence of population structure from the outset of an effort to identify phenotypically active loci, where structure or admixture is not to be used as a statistical fuel, it is generally desirable to qualify samples based on crude population stratifications such as BGA so that cases and controls can be matched and homogenized in composition. For example, it is not uncommon in the execution of case-control studies to ensure equal proportions or "racial homogeneity" within and between cases and controls.
However, for most research purposes, the subjective methods used to measure population affiliation are unsatisfactory. As currently measured using biographical questiormaires, little knowledge of population structure other than the obvious is obtained, and only basic connections between population structure and drug response can be apparent and/or controlled. Consistency is a significant problem with the self reporting of race on questionnaires, and one that the Food and Drug Administration is attempting to address during the clinical trial design process. However, using such subjective and imprecise methods of data collection, consistency can be a difficult end to achieve.

[0007] Rather than reformulating how questions are aslced on questionnaires, consistency can be better addressed by replacing the subjective nature of the exercise with objective, reproducible scientific methods. Standardization and objectivity is of paramount importance for the collection of race data because its measure can be as subjective as that of any other human attribute. The self reporting of race is not as trivial an exercise as the self reporting of gender, and many people do not lcnow their race or axe of sufficient admixture that they have trouble classifying themselves into a single group. Such a scenario is particularly common in countries such as the United States, in which numerous cultures have been combined due to immigration. For example, a woman of mainly sub-Saharan African descent, raised in Puerto, Rico, may describe herself as Hispanic. Though she socio-culturally identifies with Hispanics, however, her xenobiotic metabolism and drug target polymorphisms may more likely be associated with those shared among other sub-Saharans. By using nonanthropologic designations that describe the socio-cultural construct of society, current guidelines for considering information on race in the study design process can effect poor predictive power and false positive results. Where a person was raised and lives, and the cultural or sociological customs they observe, may have an impact on how that person responds to a S
drug or proclivity to develop a disease. Thus, non-biological metrics are required, but the evidence suggests that BGA also has an impact and, therefore, needs to be measured in a scientifically accurate and reproducible manner.
[0008] Genetic markers present in a person's DNA provide the best opportunity to reliably determine the BGA an individual, and it has long been recognized that such a means is possible. For example, Reed (Science 244:S7S-576, 1973) and Neel (Mutat.
Res.
26:319-328, 1974) referred to such markers as "private", and used them to estimate mutation rates. Reed (supra, 1973) used the term "ideal" (in reference to the utility of the maxlcers in individual ancestry estimation) to describe hypothetical genetic marker loci at which different alleles are fixed in different populations. Chakraborty et al. (Ethnic. Dis.
1:245-256, 1991) referred to variants that are found in only one population as "unique alleles", and showed how allele frequencies could be inverted to provide a likelihood estimate of population, or BGA
affiliation. The most useful "mique alleles" for the inference of BGA are those that also have large differences in allele frequency among populations (Reed, sups°a, 1973; Chakraborty et al., Gevceties 130:231-243, 1992; Stephens et al., Amef°. J. Hum.
Gevcet. SS:809-824, 1994), and that have been referred to as "population-specific alleles" (PSAs, Shriver et al., Amen. J.
Hum. Genet. 60:957-964, 1997; Parra et al., Amen. J. Hum. Genet. 63:1839-1851, 1998), but which are now referred to as "Ancestry Informative Markers" (AIMS; Shriver et al., Hum.
Gef2et. 112:387-399, 2003, Frudakis et al., J. Fof ens. Sci. 48(4) 771-782, 2003).' [0009] Within the field of forensics, statistical methods that use simple tandem repeats (STRs) to infer the highest level of ancestry in a particular individual (majority BGA using proportional ancestry notation) can be fairly robust in terms of estimating majority BGA
affiliation. Although STR tests can effectively resolve majority ancestral origin in most cases, an unacceptable number (S-10%) of classifications are ambiguous. Aside from sampling errors caused by rare alleles, and the fact that STRs were not selected from the genome for their ability to resolve population affiliation (i.e., STR allele frequency differentials are not necessarily nor optimally informative for this purpose), the major reason the high level of ambiguity lilcely is due to admixture, which is clearly a factor of the genetic variation for many human populations (Parra et al., supra, 1998, Cavalli-Sforza and Bodmer, In The genetics ofhuma~c populations (Dover Publications, NY; see pages 387-507) 1999;
Rosenberg et al., supra, 2002). For a given study design, whether using self reported information or DNA marker testing, and whether attempting to solve a pharmacogenomic or forensic problem, classifying a patient into a single group sacrifices the subtle, but not insignificant, information related to population structure and sub-structure;
for example, there is no allowance to assign a person of 50% African and SO% European affiliation into a group.
Unfortunately, markers and methods fox allowing an accurate inference as to the BGA for more than just two groups at a time for an individual have not yet been described. Thus, a need exists for robust markers useful for inferring BGA, and for methods of identifying and using such marlcers. The present invention satisfies this need, and pxovides additional advantages.
SUMMARY OF THE INVENTION

[0010] The present invention provides methods and compositions for measuring, with a desired predetermined level of confidence, within individual population structure, which, as disclosed herein, allows inferences to be drawn, for example, as to ancestry, pigmentation traits, drug responsiveness, and disease susceptibility of the individual. By way of example, the present methods and compositions were used in a forensics capacity, wherein DNA
samples obtained at the crime scenes of a serial murder/rapist in Louisiana were examined.
Based on psychological profiling, police were of the belief that the serial killer was a Caucasian male, and had tested the DNA of over 1,000 Caucasian men without finding a match. The police then turned to the inventors, who, using the compositions and methods of the invention, determined that the individual committing the crimes was African American and, more specifically, had a proportional and confidence qualified ancestry of 85% sub-Saharan African and 15% Native American. Based on this result and additional results as disclosed herein, the police were further advised that the average African American is of 20%
IndoEuropean ancestry, that greater levels of IndoEuropean ancestry correlate with lighter S1CII1 tone, and, therefore, that the person conunitting the crimes was likely an African American with average to dancer than average skin tone. Within two months of refocusing their efforts based on this information, the police arrested an African American man of average skin tone (for African Americans); DNA testing determined that he was the person whose DNA was found at the crime scenes.
j0011] Accordingly, the present invention relates to a method of inferring, with a predetermined level of confidence, a trait of an individual. Such a method can be performed, for example, by contacting a sample, which includes nucleic acid molecules of a test individual, with hybridizing oligonucleotides, wherein the hybridizing nucleotides can detect nucleotide occurrences of single nucleotide polymorphisms (SNPs) of a panel of at least about ten ancestry informative marlcers (AIMS) indicative of a population structure correlated with the trait, and wherein said contacting is performed under conditions suitable for detecting the nucleotide occurrences of the AIMS of the individual by the hybridizing oligonucleotides; and identifying, with a predetermined level of confidence, a population structure that correlates with the nucleotide occurrences of the ATMs in the individual, wherein the population structure correlates with a trait. As disclosed herein, a panel of at least about ten AIMs (e.g., 8, 9, 10, 1 l, 12, 13, 14, 15, 20, 25, 30, or more) is examined in practicing a method of the invention. Generally, the greater number of AIMS
examined, the greater the confidence level of an inference made using the method.
[0012] A trait for which an inference is made according to a method of the invention can be any trait, including a trait for which an ethnic predisposition is known or suspected to occur and a trait for which it known that no ethnic predisposition occurs or for which it is not known or unclear as to whether there is an ethnic predisposition. In one embodiment, the trait is biogeographical ancestry (BGA). In one aspect, the panel of AIMs used to examine BGA includes AIMS as set forth in SEQ ID NOS: l to 71. In another aspect, the panel includes AIMS as set forth in SEQ ID NOS:7, 21, 23, 27, 45, 54, 59, 63, and 72 to 152; in SEQ ID NOS:3, 8, 9, 11, 12, 33, 40, 59, 63, and 153 to 239; or in SEQ ID
NOS:1, 8, 11, 21, 24, 40, 172, and 240 to 331, as well as a panel containing combinations of AIMs as set forth in SEQ ID NOS:1 to 331. As disclosed herein, AIMs useful in practicing a method of the invention can, but need not, be linked to a gene linked to the trait (i.e., a gene known to be involved in the trait phenotype) and generally are not in linkage disequilibrium with the gene (or locus). For example, an AIM useful for inferring drug responsiveness of an individual g according to a method of the invention need not be linked to a gene involved in responsiveness to the drug (e.g., a drug metabolism gene or a drug transport gene such as a cytochrome P450 gene or P-glycopxotein gene). Similarly, an AIM useful for inferring a pigmentation trait of an individual according to a method of the invention need not be linked to a gene involved in pigmentation (e.g., a tyrosinase gene or a melanocortin-1 receptor gene). Thus, in one aspect, at least one (e.g., 1, 2, 3, 4, or 5) AIM of a panel is not linked to a gene involved in the trait for which an inference is being made.
[0013] Where BGA is the trait for which an inference is being made, an individual being examined can have an ancestry that includes any one ox a combination of ancestral groups, including, for example, a proportion of sub-Saharan African ancestry, Native American ancestry, IndoEuropean ancestry, East Asian ancestry, Middle Eastern ancestry, Pacif c Islander ancestry, or a combination including one or more of these ancestries.
As such, the propoz-tional ancestry of an individual can comprise one ancestry (e.g., 100%
IndoEuropean ancestry), or any proportion of two, three, four, or more ancestral groups. As such, a test individual (or individual of known proportional ancestry) can have, for example, a proportion of at least three ancestral groups, which can include proportions of sub-Saharan African ancestry and two other ancestries, ox can include proportions of sub-Saharan African and IndoEuxopean ancestral groups and a third ancestry; or Native American and IndoEuropean ancestral groups and a third ancestry; or East Asian and Native American ancestral groups and a third ancestry; or IndoEuropean and East Asian ancestral groups and a third ancestry;
or can include proportions of Native American, East Asian, and TndoEuropean ancestral groups, or of sub-Saharan African, Native American, and IndoEuropean ancestral groups, and the like.
[0014] In another embodiment, a trait of a test individual for which an inference is being made is responsiveness of the individual to a drug, particularly a therapeutic drug. As such, a method of the invention provides a tool fox realizing personalized medicine. A
drug for which an inference can be made as to whether a test individual will be responsive, in either a positive or negative manner, can be, for example, a cancer chemotherapeutic agent such as paclitaxel, or a drug such as a statin, which can be useful for maintaining or lowering cholesterol levels. In one aspect of this embodiment, AIMS of the panel of AIMS used to practice the method includes AIMS of genes other than genes known to be involved in melanin synthesis or metabolism.
[0015] In still another embodiment, a trait of a test individual for which an inference is being made is a susceptibility or predisposition of the individual to a disease. As disclosed herein, various traits are associated with population structure at a continental level, whereas other traits are associated with population structure at finer levels. As such, a method of the invention can provide an means for making an inference with respect to a trait such as disease susceptibility for diseases such as diabetes, hypertension, and cancers that are lcnoml to have an ethnic predisposition (i.e., known to occur with higher frequencies in individuals of certain ethnic/ancestral groups), as well as for disease such as such as alcoholism, or schizophreua, Parkinson's disease, and other neurological disorders, which do not (or at least are not known to) have an ethnic predisposition.
[0016] In yet another embodiment, a trait of a test individual for which an inference is being made is a pigmentation trait. The pigmentation trait can be any such trait including, for example, eye color or shade, skin color, hair color, or a combination thereof.
In one aspect of this embodiment, AIMS of the panel of AIMS used to practice the method includes AIMS of genes other than genes known to be involved in melanin synthesis or metabolism, or other aspects of pigmentation.
[0017] A method of inferring a trait of a test individual by determining a population structure that correlates with nucleotide occurrences of AIMs in the individual can further include identifying, with a predetermined level of confidence, a sub-population structure of the population structure, wherein the sub-population structure correlates with a trait. For example, a population structure of an individual can correlate to an intercontinental group with which, by inference, the individual shares ancestry, for example, IndoEuropean, and a sub-population structure can further correlate with an intracontinental group with which the individual shares IndoEuropean ancestry, for example, MediteiTanean ethnicity.

[0018] The hybridizing oligonucleotides useful in the methods of the invention can be oligonucleotide probes or oligonucleotide primers. Oligonucleotide probes useful in the present methods can hybridize to a nucleotide sequence that includes the SNP
position for an AIM, wherein the nucleotide at the position of the hybridizing oligonucleotide that corresponds to the position of the SNP for the AIM either matches or does not match the nucleotide occurrence at the SNP position. Additional oligonucleotide probes useful in the methods of the invention include oligonucleotide probes that hybridize to a polynucleotide sequence adjacent to and upstream and/or adjacent to and downstream of the SNP
position, and that can, but need not, include a nucleotide corresponding to the nucleotide position of the SNP, and wherein such a corresponding nucleotide, when present in the probe, can, but need not match the nucleotide occurrence at the SNP.
[OOI9] Oligonucleotide primers useful in the methods of the invention include oligonucleotide primers useful for a primer extension reaction, as well as oligonucleotide primers that, in combination, allow for amplification of template polynucleotide comprising the AIM. Such amplification primer pairs generally include a forward primer and a reverse primer useful for amplification of a template polynucleotide comprising an AIM
of interest.
It will be recoguzed, however, that 2, 3, 4, or more different forward primers can be used with a common reverse primer for amplification of different template polynucleotides comprising the AIM (e.g., in a multiplex reaction) and a common gene sequence (e.g., AIMs of a family of related gene sequences) or for generating amplification products of different sizes from a single template. Similarly, one common forward primer can be used with one or a plurality of different reverse primers.
[0020] Accordingly, in one embodiment, a method of the invention is performed using oligonucleotide primers. In one aspect of this embodiment, the method includes contacting the sample with the oligonucleotide primers and with a polymerase, under condition suitable for generation of a primer extension product. In such a method, the nucleotide occurrence of a SNP can be determined by detecting the presence of the primer extension product, or by sequencing the primer extension product (or a product thereof) and identifying the nucleotide at the position corresponding to the position of the SNP. In another aspect of this embodiment, the method includes contacting the sample with oligonucleotide primers that comprise amplification primer pairs and with a polymerase, under condition suitable for generation of an amplification product. In such a method, the nucleotide occurrence of a SNP can be determined by detecting the presence of the amplification product, or by sequencing the amplification product (or a product thereof) and identifying the nucleotide at the position corresponding to the position of the SNP.
[002I] The methods of the invention are particularly adaptable to being performed in a high throughput format, including in a multiplex format, thus allowing examination of a large number of AIMS and/or a large number of samples of test individuals, as well as controls, in parallel. As such, the methods can be performed using a format in which the samples being examined are arranged in an array, particularly an addressable array, e.g., on in wells in a tray or on a glass slide or silicon chip, and can be partly or fully automated using robotics. Where a multiplex platform is used, it will be recognized that the AIMS examined need not necessarily be those having the greatest delta values for the particular trait, but also can be selected to balance the delta value with the compatibility of primers in a multiplex set, for example, to select AIMS such that hybridizing oligonucleotides (e.g., amplification primer pairs) can be designed that can be used in a single reaction for examining a panel of AIMS but that do not substantially cross-hybridize with AIMS other than the target AIM
fox which the hybridizing oligonucleotides we designed.
[0022] The present invention also relates to a method of estimating, with a predetermined level of confidence, proportional ancestry of at least two ancestral groups of a test individual.
Such a method can be performed, for example, by contacting a sample, which includes nucleic acid molecules of the test individual, with hybridizing oligonucleotides that can detect nucleotide occurrences of SNPs of a panel of at least about ten AIMs that are indicative of BGA for each ancestral group examined, wherein the contacting is under conditions suitable for detecting the nucleotide occurrences of the AIMS of the test individual by the hybridizing oligonucleotides; and identifying, with a predetermined level of confidence, a population structure that correlates with, or is most likely given, the nucleotide occuxTences of the AIMS of each of the ancestral groups examined, wherein the population structure is indicative of proportional ancestry.
[0023] The proportional ancestry estimated according to a method of the invention can be a proportion of any ancestral group, including, for example, a proportion of sub-Saharan African, Native American, IndoEuropean, East Asian, Middle Eastern, or Pacific Islander ancestral group, and generally is a combination of two or more of such ancestral groups.
Thus, the proportional ancestry of a test individual can include proportions of sub-Saharan African and IndoEuropean ancestral groups (e.g., 80% sub-Saharan African and 20%
IndoEuropean; or 60% sub-Saharan African, 20% IndoEuropean, and 20% of a third ancestral group); or can include proportions of Native American and IndoEuropean ancestral groups;
East Asian and Native American ancestral groups; IndoEuropean and East Asian ancestral groups; and the like. Similarly, the proportional ancestry can include proportions of Native American, East Asian, and IndoEuropean ancestral groups; sub-Saharan African, Native American, and IndoEuropean ancestral groups; sub-Saharan African, Native American, and East Asian ancestral groups; and the lilce.
[0024] A panel of AIMS useful for estimating proportional ancestry of an individual can include AIMs as set forth in SEQ ID NOS: l to 331, for example, AIMs as set forth in SEQ
ID NOS:1 to 71, which can be useful for determining proportional ancestries including IndoEuropean, sub-Saharan African, East Asian, and Native American; or AIMS as set forth in SEQ ID NOS:7, 21, 23, 27, 45, 54, 59, 63, and 72 to 152, which can be useful for determining proportional ancestry of East Asians and sub-Saharan Africans; or in SEQ ID
NOS:3, 8, 9, 11, 12, 33, 40, 59, 63, and 153 to 239, which can be useful for determining proportional ancestry of East Asians and IndoEuropeans; or in SEQ ID NOS:l, 8, 1 l, 21, 24, 40, 172, and 240 to 331, which can be useful for determining proportional ancestry of IndoEuropeans and sub-Saharan Africans.
[0025] In one embodiment, an estimate is made wherein the proportional ancestry includes proportions of three ancestral groups. In one aspect of this embodiment, identifying a population structure that correlates with, or is most likely given, the nucleotide occurrences of the AIMS of the test individual is practiced by performing a lilcelihood determination for affiliation with each of a sub-Saharan African ancestral group, a Native American ancestral group, an IndoEuropean ancestral group, and an East Asian ancestral group;
thereafter selecting three ancestral groups having a greatest likelihood value;
determining a likelihood of all possible proportional affiliations among the three ancestral groups having the greatest likelihood value, whereby a population structure or proportional affiliation that correlates with the nucleotide occurrences of the AIMS of the test individual is identified; and identifying a single proportional combination of maximum likelihood.
[0026] In another aspect of this embodiment, identifying a population structure that correlates with, or is most likely given, the nucleotide occurrences of the AIMS is practiced by performing six two-way comparisons comprising likelihood determinations for affiliation between each group with each other group; thereafter selecting three ancestral groups having a greatest likelihood value; determining a likelihood of all possible proportional affiliations among the three ancestral groups having the greatest likelihood value, whereby a population structure or proportional affiliation that correlates with, or is most likely given, the nucleotide occurrences of the AIMs of the test individual is identified; and identifying a single proportional combination of maximum likelihood.
[0027] In still another aspect of the embodiment wherein an estimate is made wherein the proportional ancestry includes proportions of three ancestral groups, the method is practiced by performing three three-way comparisons among the groups; determining a lilcelihood of all possible proportional affiliations among the three ancestral groups having the greatest lilcelihood value, whereby a population structure or proportional affiliation that correlates with, or is most likely given, the nucleotide occurrences of the AIMS of the test individual is identified; and identifying a single proportional combination of maximum likelihood. In another aspect of this embodiment, the method can further include generating a graphical representation of the comparison of the three ancestral groups, wherein the graphical representation comprises a triangle with each ancestral group independently represented by a vertex of the triangle, and wherein the maximum likelihood value of proportional affiliation for an individual comprises a point within the triangle. If desired, the graphical representation can further include a confidence contour that indicates a level of confidence associated with estimating the proportional ancestry.
[0028] In another embodiment, an estimate is made wherein the proportional ancestry includes proportions of four ancestral groups. In various aspects of this embodiment, identifying a population structure that correlates with, or is most likely given, the nucleotide occurrences of the AIMS of the test individual is practiced by performing six two-way comparisons, or by performing three three-way comparisons, or by performing one four-way comparison among the groups; determining a lilcelihood of all possible proportional affiliations among the four ancestral groups having the greatest likelihood value, whereby a population structure or proportional affiliation that correlates with, or is most likely given, the nucleotide occurrences of the AIMs of the test individual is identified; and identifying a single proportional combination of maximum likelihood. In one aspect of this embodiment, the method can further include generating a graphical representation of the comparison of the three ancestral groups, wherein the graphical representation comprises a pyramid with each ancestral group independently represented by a vertex of the pyramid, and wherein the maximum likelihood value of proportional affiliation for an individual comprises a point witlun the pyramid. If desired, the graphical representation can further include a confidence contour comprising a sphere around the point, wherein the sphere indicates a level of confidence associated with estimating the proportional ancestry.
[0029] The method of estimating, with a predetermined level of confidence, proportional ancestry of at least two ancestral groups of a test individual by identifying a population structure indicative of the proportional ancestry can furthex include identifying a sub-population structure indicative of ethnicity associated with one of the ancestral groups for which the test individual has a proportional ancestry. According to this method, a sub-population structure of the population structure that correlates with the nucleotide occurrences of the AIMS in the test individual is identified, wherein the sub-population structure correlates with ethnicity of the test individual. Such a method of identifying a sub-population structure can be performed, for example, by identifying those chromosomes of the test individual that contain the AIMS indicative of affiliation with a BioGeographical ancestral group (where the individual is proportionally affiliated with more than one BioGeographical Ancestry group), contacting a sample including nucleic acid molecules of the test individual with second hybridizing oligonucleotides that can detect nucleotide occurrences of SNPs of a second panel of AIMS, wherein the AIMS of the second panel are informative for ethnicity within one of these groups and are present on the same chromosomes of the test individual that contain the AIMS indicative of the larger (intercontinental) ancestral group within which the ethnicity occurs; and identifying a sub-population structure that correlates with the nucleotide occurrences of the AIMS of the second panel, wherein the sub-population is indicative of etlu~icity of the ancestral group of the test individual.
[0030] According to such a method, using hybridizing oligonucleotides specific for the first panel of AIMS (e.g., AIMs of the 71 exemplified AIMS; SEQ ID NOS:1 to 71), a test individual can be determined to be 60% IndoEuropean (IE) and 40% East Asian.
In such a case, only a fraction of the total possible AIMS that can be indicative of the IE ancestral group will have been positive (if all were positive, the individual would have been 100% IE) and, therefore, only some of the individuals chromosomes or chromosomal regions will be of IndoEuropean origin. The chromosomes of the individual containing the positive AIMS for IE are then identified, and second hybridizing oligonucleotides specific for a second panel of AIMS are selected (e.g., from a group of 1000 or so AIMS that cover all 23 pairs of human chromosomes), wherein the AIMS of the second panel are limited to those that are highly variable in allele frequencies between IE ethnic groups and, therefore, indicative of IE
ethnicity, and also are present on the chromosomes for which the first panel AIMS were IE
positive. A sub-population structure that correlates with the nucleotide occurrences of the AIMS of the second panel is then identified, thus indicating an ethnicity with respect to the IE
ancestral group of the test individual, for example, that the IE ancestral group derives from a Northern European, a Mediterranean, a Middle Eastern, or a South Asian Indian ethnicity.
As such, the method provides a means to identify the ethnic origin of particular chromosomes (e.g., a Mediterranean origin of chromosomes previously determined to be of IndoEuropean origin) that contain AIMS that correlate with a population structure indicative of IndoEuropean BioGeographical Ancestry, and further contain AIMS that correlate more specifically with a sub-population structure indicative of Mediterranean ethnicity.
[0031] In another embodiment, the method of estimating proportional ancestry of a test individual can~include generating an ancestral map of the world, wherein locations of populations having a proportional ancestry corresponding to the proportional ancestry of the test individual are indicated on the ancestral map. As such, the method can supplement genealogical information. Fox example, the method can further include overlaying the ancestral map with a genealogical map, wherein the genealogical map indicates locations of populations having geopolitical relevance with respect to the test individual, and statistically combining the information of the ancestral map and genealogical map to obtain a most likely estimate of family history of the test individual [0032] Identifying a population structure that coiTelates with, or is most likely given, the nucleotide occurrences of the AIMs, according to a method of the invention, can be performed by comparing the nucleotide occurrences of the AIMS of the test individual with known proportional ancestries corresponding to nucleotide occurrences of AIMS
indicative of BGA. The lcnown proportional ancestries corresponding to nucleotide occurrences of AIMS
indicative of BGA can be contained in a table or other list, and the nucleotide occurrences of the test individual can be compared to the table or list visually, or can be contained in a database, and the comparison can be made electronically, for example, using a computer.
Further, each of the known proportional ancestries corresponding to nucleotide occurrences of AIMS indicative of BGA can be associated with a photograph of a person from whom the known proportional ancestry was determined, thus providing a means to further infer physical characteristics of a test individual. In one aspect, the photograph is a digital photograph, which comprises digital information that can be contained in a database that can further contain a plurality of such digital information of digital photographs, each of which is associated with a lcnown proportional ancestry corresponding to nucleotide occurrences of AIMS indicative of BGA of the person in the photographs.
[0033] In another aspect, a method of the invention can further include identifying a photograph of a person having a proportional ancestry corresponding to the proportional ancestry of the test individual. Such identifying can be done by manually looking through one or more files of photographs, wherein the photographs are organized, for example, according to the nucleotide occurrences of ATMs of the person in the photograph. Identifying the photograph also can be performed by scanning a database comprising a plurality of files, each file containing digital information corresponding to a digital photograph of a person having a known proportional ancestry, and identifying at least one photograph of a person having nucleotide occurrences of ATMs indicative of BGA that correspond to the nucleotide occurrences of AIMS indicative of BGA of the test individual.
[0034] Accordingly, the present invention also relates to an article of manufacture, which is at least one photograph of a person having a known proportional ancestry corresponding to a population structure comprising nucleotide occurrences of AIMS indicative of BGA, as well as to a plurality of such articles, each article of the plurality comprising one (or more) photographs) of a person having a known proportional ancestry corresponding to a population structure comprising nucleotide occurrences of AIMS indicative of BGA. The article can be contained in a file, or a plurality of the articles can be contained in a filed, for example, a file containing a plurality of photographs of different persons, wherein the some or all of the persons have the same or different known proportional ancestries that correspond to a population structure comprising nucleotide occurrences of AIMS indicative of BGA.
[0035] Accordingly, a plurality of such articles is provided, as is a plurality of files, each file of which can contain one or more articles, i.e.,, photographs, which can be of one or more persons having the same or different l~nown proportional ancestries that correspond to a population structure comprising nucleotide occurrences of AIMS indicative of BGA. For example, different files of the plurality each can contain one (or more) photographs) of one person having a lcnovnnl proportional ancestry corresponding to a population structure comprising nucleotide occurrences of AIMS indicative of BGA. Different files of the plurality also can contain photographs of two or more different persons, each of whom has the same ox substantially the same proportional ancestry corresponding to a population structure comprising nucleotide occurrences of AIMS indicative of BGA. As such, a plurality of files can contain files, each of which contains one or more photographs of one or more persons, and when containing one or more photographs of two or more different persons, the different persons can have the same or different known proportional ancestries.
(0036] In one embodiment, the article of manufacture, i.e., the photograph of a person having a known proportional ancestry corresponding to a population structure comprising nucleotide occurrences of AIMS indicative of BGA, is a digital photograph, which comprises digital information. As such, the digital information of the digital photograph, or of a plurality of digital photograph articles of manufacture of the invention can be contained in a database. As such, the present invention further provides a plurality of the articles of manufactures, including at least two digital photographs each of which comprises digital information. In one aspect of this embodiment, the digital information for one or a plurality of the articles is contained in a database, which can be contained in any medium suitable for containing such a database, including, for example, computer hardware or software, a magnetic tape, or a computer disc such as floppy disc, CD, or DVD. As such, the database can be accessed through a computer, which can contain the database therein, can accept a medium containing the database, or can access the database through a wired or wireless networlc, e.g., an intranet or Internet.
[0037] The present invention also relates a lcit, which contains a plurality of hybridizing oligonucleotides, each hybridizing oligonucleotide including at least fifteen contiguous nucleotides of a polynucleotide as set forth in SEQ ID NOS: l to 331, or a polynucleotide complementary thereto, and the plurality including at least five of such oligonucleotides, each based on different polynucleotides as set forth in SEQ ID NOS:1 to 331. In one embodiment, the hybridizing oligonucleotides that include at least fifteen contiguous nucleotides of at least five polynucleotides as set forth in SEQ ID NOS:1 to 71, or polynucleotides complementary to any of SEQ ID NOS:1 to 71.
[0038] The hybridizing polynucleotides of a lcit of the invention can include probes, which are useful for detecting a particular AIM, including a particular nucleotide occurrence at the SNP position or DIP (deletioWinsertion polymorphism) position of the AIM; can include primers, including primers useful for a primer extension reaction and primer pairs useful for a nucleic acid amplification reaction; or can include combinations of such probes and primers.

In one embodiment, a hybridizing oligonucleotide of the plurality includes a nucleotide corresponding to nucleotide position of the AIM (e.g., nucleotide 50 of any of SEQ ID
NOS:1 to 34 and most others, nucleotide 56 of SEQ ID N0:35, nucleotide 44 of SEQ ID
NO:50, or nucleotide 26 of SEQ ID N0:56), or to a nucleotide sequence complementary thereto, such a hybridizing oligonucleotide being useful as a probe to identify the presence or absence of a particular nucleotide occurrence at the SNP position of the AIM.
[0039] In another embodiment, the kit contains at least one pair of hybridizing oligonucleotides useful for detecting the nucleotide occurrences) at the SNP
(or DIP) position of an AIM. In one aspect of this embodiment, a pair of hybridizing oligonucleotides includes one oligonucleotide that hybridizes upstream and adjacent to the SNP
position of an AIM and a second oligonucleotide that hybridizes downstream of and adjacent to the SNP (or DIP) position of the AIM, wherein one or the other of the pair further contains a nucleotide complementary to a nucleotide occurrence suspected of being at the SNP (or DIP) position of the AIM (i.e., one of the polymorphic nucleotides), such a pair of hybridizing oligonucleotides being useful in an oligonucleotide ligation assay. In another aspect of this embodiment, a pair of hybridizing oligonucleotides includes an amplification primer pair, including a forward primer and a reverse primer, such a pair of hybridizing oligonucleotides being useful for amplifying a portion of polynucleotide that includes the SNP
(or DIP) position of the AIM.
[004.0) A lcit of the invention can fiu~ther contain additional reagents useful for practicing a method of the invention. As such, the lcit can contain one or more polynucleotides comprising an AIM, including, for example, a polynucleotide containing an AIM
for which a hybridizing oligonucleotide or pair of hybridizing oligonucleotides of the lit is designed to detect, such polynucleotide(s) being useful as controls. Further, hybridizing oligonucleotides of the lcit can be detectably labeled, or the lit can contain reagents useful for detectably labeling one or more of the hybridizing oligonucleotides of the kit, including different detectable labels that can be used to differentially label the hybridizing oligonucleotides; such a lcit can further include reagents for linking the label to hybridizing oligonucleotides, or for detecting the labeled oligonucleotide, or the like. A lit of the invention also can contain, for example, a polymerase, particularly where hybridizing oligonucleotides of the kit include primers or amplification primer pairs; or a ligase, where the lit contains hybridizing oligonucleotides useful for an oligonucleotide ligation assay. In addition, the kit can contain appropriate buffers, deoxyribonucleotide triphosphates, etc., depending, for example, on the particular hybridizing oligonucleotides contained in the lcit and the purpose for which the kit is being provided.
BRIEF DESCRIPTION OF THE DRAWINGS
[0041] Figure 1 provides a diagram indicating the fashion in which chromosomal segments are shuffled by recombination over time in an admixed population.
Initially, the parental populations have chromosomal segments that are continuous with respect to AIMS
along the segment. In the first filial (F1) generation all persons have one complete chromosomal segment from each parental population. In the F2 generation, many more combinations are possible. The relative likelihood of the non-recombinant vs.
the recombinant genotypes shown in F2 is dependent on the size of the chromosomal segment.
Segments of the order of the size of human chromosomes will average several recombination events in a single meiosis (one recombination is equally likely every 50 cM of genetic distance). F3 shows an example of a likely genotype for a person with two parents from the F2 generation. F(N) x F1 diagrams a genotype of a person with one F(N) parent and one F1 parent; and F(N) x F2 diagrams a genotype of a person with one F(N) parent and one F2 parent.
[0042] Figures 2A and 2B show triangle graphs generated using the algorithm described in Example 6 (see, also, Table 12). NAM, Native American; AFR, sub-Saharan African;
EUR, IndoEuropean.
[0043] Figure 2A illustrates extension of a line from the NAM vertex to the opposite leg of the triangle, wherein the opposite leg represents 0% Native American ancestry. A circle is shown at the position of the estimated proportional ancestry (see Figure 2B), with the hatch mark on the line indicated the percent of Native American ancestry (approximately 15%).

[0044] Figure 2B shows additional lines drawn from the AFR and EUR vertices.
The position on each line corresponding to the position of the circle represents the proportion of each respective ancestry; i.e., 15% Native American, 60% IndoEuropean, and 25%
African.
[0045] Figure 3 shows a triangle plot depicting one approach to illustrate the value and precision of individual ancestry estimates. Typical distributions of three populations are shown (European Americans: filled squares; African-Americans: open triangles;
and an African/Native American population: open circles). Also shown is a single individual with likelihood intervals represented as concentric rings surrounding the point estimate (filled circle). Like a topological map, each concentric ring represents a decrease in the likelihood by 1 log unit (10 times less likely). In this example, the individual has a likelihood interval space that is symmetrical and circular. Interval spaces will take many shapes depending on the admixture proportions of the subject in question and the allele frequencies of the markers that have been typed.
[0046] Figure 4 provides a triangular plot showing average admixture estimates for three African-American samples (filled circles: WASH-Washington DC, AFCAR-AfroCaribbeans, and BOG-Bogalusa), a European-American sample (open circle: SCO-State College), and a Spanish-American sample (open diamond: SLV-San Luis Valley CO). In parenthesis is shown the average African (AFR), IndoEuropean (EUR), and Native American (NAM) genetic contribution to each sample.
[0047] Figures SA and SB show the genetic structure in US resident populations.
[0048] Figure SA shows the percentage of unlinked AIMS showing significant association.
Expected values are based on a 5% significance level. Values for the Washington DC sample are based on 33 AIMS, for San Luis Valley CO on 19 AIMS, and for State College PA on 34 AIMs.
[0049] Figure SB shows the correlation between individual ancestry estimates based on independent subsets of informative markers. Average correlation is based on 100 replicates.

The total number of markers is the same as for Figure SA. The corresponding p values are indicated at the bottom of the graph.
[0050] Figures 6A and 6B show the triangle plots for a father (Figure 6A) and mother (Figure 6B).
[005I] Figures 7A to 7C show the triangle plots for each of three children of the father and mother represented in Figure 6.
[0052] Figure 8 shows the distribution of AIMS in the genome (chrom. number, chromosome number).
[0053] Figures 9A and 9B demonstrate the robustness of BGA admixture proportion analysis using AIMS (see Example 2). The confidence (contour lines) of the maximum likelihood estimate (MLE; point) is predictably affected by the elimination of AIMs informative for a particular pair-wise comparison. The first contour line extending from the MLE defines the triangle plot space within which the lilcelihood is 2 times lower than that of the MLE, and the second contour line defines the space in which the likelihood is 5 times lower than the MLE.
[0054] Figure 9A shows the MLE and confidence contours obtained using 71 AIMS;
actual percentages are indicated.
[0055] Figure 9B shows the results obtained after eliminating those AIMs used to obtain the results shown in Figure 9A from the analysis that are informative for East Asian Native American distinction. The MLE is relatively unaffected, and the confidence contours along the East Asian-Indo European (European) and Native American-European axes remain undistorted, but the confidence contours are distorted along the East Asian-Native American axis.
[0056] Figure 10 shows the BGA admixture proportions determined for each of eight individuals of a family pedigree. Circles represent females, squares males and the BGA
affiliation for each individual is shown as a fraction where the numerator represents IndoEuropean BGA and the denominator represents Native American BGA. None of the individuals harbored sub-Saharan African or East Asian BGA except as indicated by the asterisk ('~), which indicates that the individual was determined to be of 4%
East Asian BGA.
(0057] Figure 11 shows a family tree demonstrating how a Chinese great grandparent in an otherwise IndoEuropean family tree can produce a grandchild with IndoEuropean/East Asian ancestry. The individuals that are 100% East Asian (Chinese) are shown with shading;
the admixture results for the male (square) at the bottom of the pedigree (short arrow) are of interest. The grandparent indicated by the long arrow is about a 50%/50% East Asian/IndoEuropean mix, and her daughter, the subject's mother, is expected to be a 25%/75% East Asian/IndoEuropean mix (see Example 3).
[0058] Figure 12 shows the distribution of all SNPs available for genotyping by chromosomal arm for a group of patients treated for elevated cholesterol levels.
[0059] Figure 13 shows the distribution of SNPs in Caucasian individuals taking Lipitor~
(n = 180) for whom response was known in terms of cholesterol (lip TC), low density lipoprotein (lip LDL), liver transaxninase AST-SGOT (lip SGOT) and ALT-GPT
(lip GPT) measurements. SNPs with delta values of significance (>0.20) among the various trait classes were selected. For example, in about 70% of patients, LipitorTM causes a decrease in LDL.
For any given SNP, the delta value (8) is the difference in minor allele frequency among those individuals for whom LDL decreased by at least 20% versus those for whom LDL did not change.
[0060] Figure 14 shows a similar analysis as for Figure 13, except that response is measured following treatment with ZocorTM (n = 150), and only total cholesterol (zoc TC) and LDL (zoc LDL) were examined.
[0061] Figure 15 shows a distribution of SNPs (b > 0.11) among chromosome for 1,000 individuals of known eye color.
[0062] Figure 16 shows a distribution of SNPs (b > 0.11) among chromosome for 1,000 individuals of known hair eye color.

DETAILED DESCRIPTION OF THE INVENTION
[0063] The present invention is based on the identification of ancestry informative markers (AIMS) useful for inferring a level of population structure of an individual, which, in turn, allows an inference as to various traits of the individual. Further, the AIMs of the present invention are demonstrated to correlate with a trait, regardless of whether the marker is in liu~age disequilibrium with a gene or locus known to be involved in the trait. As such, the AIMs of the present invention are distinguishable from previously described marlcers, which only were considered useful if they were linked with a trait, i.e., if the marker was physically close to a gene known to be involved in the trait as characterized, for example, in having a low cross-over percentage with respect to gene (or locus) known to be involved in (or associated with) the trait. In contrast, there is no requirement that the marlcers (AIMS) useful in the present methods be in linkage disequilibrium with a gene/trait and, in fact, AIMS
that are disclosed herein as correlating with a trait can be Located on different chromosomes fiom each other and from a gene/locus known to be associated with the trait.
[0064] AIMS are genetic loci that show alleles with high frequency differences between populations. AIMS are exemplified herein generally by single nucleotide polymorphisms (SNPs; see, e.g., SEQ ID N0:1), as well as by deletion/insertiori polymorphisms (DIPS; see, e.g., SEQ ID NO:363). As disclosed herein, AIMS can be used to estimate BioGeographical Ancestry (BGA) of an individual or collection of individuals at the population level (in terms of races), at the sub-population level (in terms of ethnicities), and at the micro-group level (in terms of familial lines within ethnic groups), as well as at a practical, phenotypically qualified level (e.g., cases and controls). Such ancestry estimates at the subgroup and individual level can be directly instructive regarding the genetics of phenotypes that are different qualitatively or in frequency between populations, including, for example, the likelihood that an individual will respond to a particular medication or the propensity of an individual to develop a disease.
Ancestry estimates also can provide a compelling foundation for the use of Admixture Mapping (AM) methods to identify the genes underlying these traits.
[0065] As exemplified herein, a panel of 71 AIMs (SEQ ID NOS:1 to 71) was identified from an examination of over 800 candidate AIMS (see, also, SEQ ID NOS:72 to 331), and methods were developed to examine these AIMS as a means to obtain accurate estimates of proportional ancestry. The methods and markers of the invention have been validated in studies using skin pigmentation as a model phenotype (see, also, Intl. Publ.
No. WO 02/097047 (PCT/US02/16789), which is incorporated herein by reference).
Initial markers were genotyped in two population samples with primarily African ancestry, African Americans from Washington D.C. and an African Caribbean sample from England, and in a sample of European Americans from Pennsylvania (see Example 1). In the two African population samples, very strong correlations were obsexved between estimates of individual ancestry and shin pigmentation as measured by reflectometry (R2 = 0.21, p <
0.0001 for the African-American sample and R2 = 0.16, p < 0.0001 fox the British African-Caribbean sample). These correlations confirmed the validity of the ancestry estimates and also indicated the high level of population structure related to admixture, which characterizes these populations and is detectable using other tests to identify genetic structure. These results demonstrate that an estimate of an individual's ancestry can be made based on a DNA
analysis using a relatively small number of well defined genetic markers (AIMS).
[0066] The methods and genetic markers disclosed herein provide tools for several distinct purposes, including, for example, 1) for the estimation of ancestry pxoportions in individuals from their DNA; 2) for the estimation of genetic structure for the control of study designs commonly used for genetic research; 3) for the construction of physical profiles through the inference of characteristics related to ancestry, which may have implications in forensic investigations; 4) for the identification of disease predisposition, referred to as "Mapping by Ancestry Linkage Disequilibrimn" (MALD); and 5) for predicting a significant portion of an individual patient's response to prescription and over-the-counter medications. AS such, the present invention provides, for example, 1) statistical methods for the determination of ancestral proportions from genetic sequences within individuals and examples of use;
2) several hundred AIMS culled from the publicly available single nucleotide polymorphism (SNP) database and identified using statistical methods as useful for the determination of ancestral proportions within individuals or study groups; 3) several hundred AIMS that are demonstrated as useful for the determination of ancestral proportions within individuals or study groups; and 4) software programs that can be used for the determination of ancestral proportions within individuals or study groups [0067] Previously, efforts have been made to control the two sources of population structure, including sampling effects and natural human demography, which were believed to confomld efforts to identify markers of genes associated with pal-ticulax traits. However, as disclosed herein, population structure is reflective of human demography, and marlcers that correlate with a trait value are useful as reporters of structure that correlate with trait value (rather than markers in LD with phenotypically active loci), and, therefore, provide a valuable tool that enables accurate classification in a cost-effective and practical manner. Alleles associated with a trait due to population structure are not linked to phenotypically active loci, but are merely correlated with trait value because they are enriched for in branches of the human family tree for which the trait value is more common. As disclosed herein, the distribution of trait values among the various branches of the human family tree are such that accurate classification can be obtained only through an appreciation of that structure, rather than a full understanding of the biological mechanism of the trait, and, as a result, markers that were considered false positives when considered with respect to their use for identifying phenotypically active loci, in fact, can enable accurate classification analysis; i.e., they are true positives provided the structure from which they were derived is reflective of human demography rather than sampling effects. The present methods are based on correlation between markers and BGA, where BGA is itself on some level of complexity correlated with a trait value, not Iinlcage or linlcage disequilibrium.
[0068] Accordingly, the present invention provides a method of inferring, with a predetermined level of confidence, a trait of an individual. In one embodiment, a method of the invention is performed by contacting a nucleic acid sample of a test individual with hybridizing oligonucleotides that can detect nucleotide occurrences of single nucleotide polymorphisms (SNPs) of a panel of at least about ten AIMS; and identifying, with a predetermined level of confidence, a population structure that correlates with, or is most likely given, the nucleotide occurrences of the AIMS in the individual, wherein the population structure correlates with a trait. The panel of AIMS are selected on their delta value (see below) and, where relevant, based on the particular platform used to perform the method, and are indicative of a population structure correlated with the trait. AIMs are exemplified herein by the polynucleotides set forth as SEQ ID NOS: l to 331, wherein the SNP
position generally is at nucleotide position 50 (but see, e.g., SEQ ID N0:35, nucleotide 56; SEQ ID
NO:51, position 48; SEQ ID N0:56, position 26).
[0069] A test individual for whom a trait is to be inferred can be any individual for whom it is desired to infer a trait, and generally is a human. However, the methods of the invention also can be used for inferring traits of other mammals, including, for example, domestic animals such as cats, dogs, or horses; farm animals such as cattle, sheep, pigs, or goats; or other animals. The trait to be examined can be any trait of interest, including, as exemplified herein, proportional ancestry (BGA); hair, skin or iris pigmentation; or drug responsiveness.
[0070] The methods of the invention are particularly useful because they allow for an inference to be made of a desired trait with a predetermined level of confidence. As used herein, reference to a "predetermined level of confidence" means that an inference or estimate of the invention is made using statistical methods that provide a confidence interval to be determined about a mean or a maximum likelihood value. In addition to determining the maximum likelihood value of within-individual or within-sample structure, other similarly likely values can also be determined and these can be combined to define the x-fold likelihood confidence intervals, where x is any number such as 2, 5 or 10. For example, all of the structure results corresponding to a likelihood value 10 times lower than the Maximum Likelihood Value can be plotted or listed to define the 10-fold likelihood confidence interval.
As for any statistical test, an assay of the invention is designed such that performance of the test results in a value having a desired confidence level. As disclosed herein, a method of the invention can be performed such that the result has a predetermined level of confidence by varying the number of AIMS examined with respect to a trait. For example, use of a certain panel of ten AIMS will allow an inference to be made as to whether an individual has a particular trait, e.g., responsiveness to LipitorTM, with a certain level of confidence, whereas use of a panel of twenty AIMS, which can, but need not be partially overlapping with the panel of ten AIMS, will allow the same inference to be made, but with a higher level of confidence. Similarly, use of two panels of ten AIMS each can allow an inference to be made that an individual has, for example, 80% IndoEuropean ancestry and 20% East Asian ancestry (with an error, e.g., of ~ 10%), whereas the use of two panels of twenty AIMS each can allow the same inference, but with an error, e.g., of + S%.
[0071] A sample useful for practicing a method of the invention can be any biological sample of a test individual that contains nucleic acid molecules, including portions of the gene sequences contaiung AIMS that axe to be examined or, wherein the polymorphism of an AIM results in an amino acid change in an encoded polypeptide, any biological sample that contains the encoded polypeptides. As such, the sample can be a cell, tissue or organ sample, or can be a sample of a biological fluid such as semen,. saliva, blood, cerebrospinal fluid, and the like.
[0072] A nucleic acid sample useful for practicing a method of the invention will depend, in part, on whether the SNPs to be identified are in coding regions or in non-coding regions.
Where one or more SNPs is present in a non-coding region of a gene, the nucleic acid sample generally is a deoxyribonucleic acid (DNA) sample, particularly genomic DNA or an amplification product thereof. However, where the AIM is contained within a transcribed sequence, e.g., rDNA, microsatellite DNA, or heteronuclear ribonucleic acid (RNA), which includes unspliced mRNA precursor RNA molecules including non-coding RNA
sequence, an RNA sample can be used and examined directly, or a cDNA or amplification product thereof can be examined according to the present methods. Where one or more SNPs is present in a coding region of a gene, the nucleic acid sample can be DNA or RNA, or products derived therefrom, for example, amplification products. Furthermore, while the methods of the invention are exemplified with respect to a nucleic acid sample, it will be recognized that particular SNPs, when present in coding regions of a gene, can result in polypeptides containing different amino acids at the positions corresponding to the SNPs due to non-degenerate codon changes. As such, in one aspect, the methods of the invention are practiced using a sample containing polypeptides of the subject.
[0073] A method of the invention is performed by contacting the sample and hybridizing oligonucleotides under conditions suitable for detecting the nucleotide occurrences of the AIMS of the individual by the hybridizing oligonucleotides. Further, in aspects of the methods of the invention, the sample can be contacted with second hybridizing oligonucleotides, fox example, to determine a sub-population structure. It should be recognized that the temp "second", when used in reference to hybridizing oligonucleotides (or to a panel of AIMs), is used for convenience of discussion so as to allow a clear distinction, e.g., of steps for performing a method. In this respect, it should be further recognized that one or more hybridizing oligonucleotides used, e.g., to determine a population structure, also can be included among the second hybridizing oligonucleotides.
[0074] Conditions suitable for detecting the nucleotide occurrences of AIMS
will vary depending on the sequences of the hybridizing oligonucleotides, including their length and complementarity, as well as on the particular assay being used and, for example, whether the assay is being performed as a multiplex assay. The hybridizing oligonucleotides, which axe at least 15 nucleotides in length, can contain deoxyribonucleotides or ribonucleotides, which are linked together by a phosphodiester bond, and can be single stranded or double stranded, though they generally axe used in a single stranded form. Such hybridizing oligonucleotides can be prepared using methods of chemical synthesis or by enzymatic methods such as by the polymerase chain reaction (PCR).
[0075] The hybridizing oligonucleotides, or other polynucleotides useful in a methods or contained in a lcit of the invention also can contain nucleoside or nucleotide analogs, and can have a backbone bond other than a phosphodiester bond, such oligonucleotides providing certain advantages such as having increased stability or more desirable hybridization properties. Nucleotide analogs are well known in the art and commercially available, as are polynucleotides containing such nucleotide analogs (Lin et al., Nucl. Acids Res.
22:5220-5234, 1994; Jellinelc et al., Biochemistry 34:11363-11372, 1995;
Pagratis et al., Natatre Bioteehnol. 15:68-73, 1997, each of which is incorporated herein by reference). The covalent bond also can be any of numerous other bonds, including a thiodiester bond, a phosphorotluoate bond, a peptide-like bond or any other bond laiown to those in the art as useful for linlcing nucleotides to produce synthetic oligonucleotides (see, for example, Tam et al., Nucl. Acids Res. 22:977-986, 1994; Ecker and Crooke, BioTech~cology 13:351360, 1995, each of which is incorporated herein by reference). The incorporation of non-naturally occurring nucleotide analogs or bonds linking the nucleotides or analogs can be particulaxly useful where the oligonucleotide is to be exposed to an environment that can contain a nucleolytic activity, including, for example, a tissue culture medium or sample comprising a cell extract because the modified oligonucleotides can be less susceptible to degradation.
[0076] Generally, the hybridizing oligonucleotides useful for purposes of the present invention are at least about 15 bases in length, which is sufficient to permit the oligonucleotide to selectively hybridize to a target polynucleotide comprising the AIM, and can be at least about 18 nucleotides or 21 nucleotides or 25 nucleotides or more in length.
The term "selective hybridization" or "selectively hybridize" refers to hybridization under moderately stringent or highly stringent physiological conditions, which can distinguish related nucleotide sequences from unrelated nucleotide sequences. In nucleic acid hybridization reactions, the conditions used to achieve a particular level of stringency are known to vary, depending on the nature of the nucleic acids being hybridized, including, for example, the length, degree of complementarity, nucleotide sequence composition (e.g., relative GC:AT content), and nucleic acid type, i.e., whether the oligonucleotide or the target nucleic acid sequence is DNA or RNA. An additional consideration is whether one of the nucleic acids is immobilized, for example, on a filter, bead, chip, or other solid matrix.
[0077] Methods for selecting appropriate stringency conditions can be determined empirically or estimated using various formulas, and are well known in the art (see, for example, Sambrook et al., supra, 1989). An example of progressively higher stringency conditions is as follows: 2X SSC/0.1% SDS at about room temperature (hybridization conditions); 0.2X SSC/0.1% SDS at about room temperature (low stringency conditions);
0.2X SSC/0.1% SDS at about 42°C (moderate stringency conditions); and O.1X SSC at about 68°C (high stringency conditions). Washing can be carried out using only one of these conditions, for example, high stringency conditions, or each of the conditions can be used, for example, for 10 to 15 minutes each, in the order listed above, repeating any or all of the steps listed. As such, final conditions will vary, depending on the particular hybridization reaction involved, and can be determined empirically. It should be recognized that a variety of conditions can be utilized to provide selective hybridization conditions. For example, when a multiplex assay is to be performed using a plurality of different hybridizing oligonucleotides specific for different AIMs of a panel, the conditions (as well as the AIMs/hybridizing oligonucleotides) can be selected such that selective hybridization occurs for all of the hybridizing oligonucleotides in the reaction.
[0078] In various embodiments, it can be useful to detectably label a polynucleotide or hybridizing oligonucleotide. Detectable labeling of a polynucleotide is well known in the art and includes, for example, the use of detectable labels such as chemiluminescent labels, radionuclides, enzymes, haptens such as digoxygenin and biotin, fluorophores, and unique oligonucleotide sequences. For example, PCR products can be performed, wherein one primer is biotinylated and the other primer contains digoxygenin. The amplification products can then be bound to a streptavidin plate, washed, reacted with an enzyme-conjugated antibody to digoxygenin, and developed with a chromogenic, fluorogenic, or chemiluminescent substrate for the enzyme. Alternatively, a radioactive method can be used to detect generated amplification products, for example, by including a radiolabeled deoxynucleoside triphosphate into the amplification reaction, then blotting the amplification products onto DEAE paper for detection. In addition, if one primer is biotinylated, then streptavidin-coated scintillation proximity assay plates can be used to measure the PCR
products. Additional methods of detection can use a chemiluminescent label, for example, a lanthanide chelate such as used in the DELFIA~ assay (Pall Corp.), a fluorescent label, or an electrochemiluminescent label such as ruthenium tris-bipyridyl (ORI-GEN).
[0079] Methods for detecting a nucleotide occurrence at a SNP or DIP position of an AIM
can utilize one or more oligonucleotide probes or primers, including, for example, an amplification primer pair, that selectively hybridize to a taxget polynucleotide spanning the AIM. Oligonucleotide probes useful in practicing a method of the invention can include, for example, an oligonucleotide that is complementary to and spans a portion of the taxget polynucleotide, including the position of the SNP (or DIP), wherein the presence of a specific nucleotide at the position of the SNP is detected by the presence or absence of selective hybridization of the probe. Such a method can further include contacting the target polynucleotide and hybxidized oligonucleotide with an endonuclease, and detecting the presence or absence of a cleavage product of the probe, depending on whether the nucleotide occurrence at the SNP site is complementary to the corresponding nucleotide of the probe. A

pair of probes that specifically hybridize upstream and adjacent and downstream and adjacent to the site of the SNP, wherein one of the probes includes a nucleotide complementary to a nucleotide occurrence of the SNP, also can be used in an oligonucleotide ligation assay, wherein the presence or absence of a ligation product is indicative of the nucleotide occurrence at the SNP site. An oligonucleotide also can be useful as a primer, for example, for a primer extension reaction, wherein the product (or absence of a product) of the extension reaction is indicative of the nucleotide occurrence. In addition, a primer pair useful for amplifying a portion of the target polynucleotide including the SNP or DIP
site can be useful, wherein the amplification product is examined to determine the nucleotide occurrence at the SNP site or to determine whether there is an insertion or a deletion at the DIP site.
[0080] Numerous methods are la~own for determining a nucleotide occurrence at a particular position in a polynucleotide (i.e., of a SNP or DIP). Such methods can utilize one or more oligonucleotide probes or primers, including, for example, an amplification primer pair, that selectively hybridize to a target polynucleotide, which contains one or more SNP
positions. Hybridizing oligonucleotide useful in practicing a method of the invention can include, for example, an oligonucleotide that is complementary to and spans a portion of the target polynucleotide, including the position of the SNP or DIP (including whether the DIP
has a deletion or insertion), wherein the presence of a specific nucleotide at the SNP site or the presence of a deletion or insertion at the DIP site is detected by the presence or absence of selective hybridization of the oligonucleotide probe. Such a method can further include contacting the target polynucleotide and hybridized oligonucleotide with an endonuclease, and detecting the presence or absence of a cleavage product of the pxobe, depending on whether the nucleotide occurrence at the SNP site is complementary to the corresponding nucleotide of the probe.
[0081] An oligonucleotide ligation assay also can be used to identify a nucleotide occurrence at a SNP site, wherein a pair of probes that selectively hybridize upstream and adjacent to and downstream and adjacent to the site of the SNP, and wherein one of the probes includes a terminal nucleotide complementary to a nucleotide occurrence of the SNP.
Where the terminal nucleotide of the probe is complementary to the nucleotide occurrence, selective hybridization includes the terminal nucleotide such that, in the presence of a ligase, the upstream and downstream oligonucleotides are ligated. As such, the presence or absence of a ligation product is indicative of the nucleotide occurrence at the SNP
site.
[0082] A hybridizing oligonucleotide also can be useful as a primer, for example, for a primer extension reaction, wherein the product (or absence of a product) of the extension reaction is indicative of the nucleotide occurrence at a SNP site or an insertion or deletion at a DIP site. In addition, a primer pair useful for amplifying a portion of the target polynucleotide including the SNP or DIP site can be useful, wherein the amplification product is examined to determine the nucleotide occurrence at the SNP site or the presence of a deletion or an insertion at the DIP site. Particularly useful methods include those that are readily adaptable to a high throughput format, to a multiplex format, or to both.
[0083] Conditions that allow generation of an amplification product in a sample in which an amplif canon reaction is being performed are such that the reaction contains the necessary components for the amplification reaction to occur. Such conditions include, for example, appropriate buffer capacity and pH, salt concentration, metal ion concentration if necessary for the particular polymerase, appropriate temperatures that allow for selective hybridization of the primer or primer pair to the template target polynucleotide, as well as appropriate cycling of temperatures that permit polymerase activity and melting of a primer or primer extension or amplification product from the template or, where relevant, from forming a secondary structure such as a stem-loop structure. Such conditions and methods for selecting such conditions are routine and well known in the art (see, for example, Innis et al., "PCR
Strategies" (Academic Press 1995); Ausubel et al., "Short Protocols in Molecular Biology"
4th Edition (John Wiley and Sons, 1999), each of which is incorporated herein by reference).
[0084] A primer extension or amplification product can be detected directly or indirectly andlor can be sequenced using various methods lcnown in the art. Amplification products that span a SNP site can be sequenced using traditional sequence methodologies, including, for example, the dideoxy-mediated chain termination method (Sanger et al., J.
Molec. Biol.
94:441, 1975; Prober et al. Science 238:336-340, 1987) or the chemical degradation method (Maxam et al., Pooc. Natl. Acad. Sci. USA 74:560, 1977) to determine the nucleotide occurrence at the SNP loci.

[0085] The nucleotide occurrence at a SNP site also can be determined using a microsequencing method, wherein the identity of only a single nucleotide is determined at a predetermined site (U.S. Pat. No. 6,294,336). Microsequencing methods include the Genetic Bit Analysis method (WO 92/15712). Additional, primer-guided, nucleotide incorporation procedures fox assaying polymorpluc sites in DNA have also been described (Komher et al., Nucl. Acids. Res. 17:7779-7784, 1989; Solcolov, Nucl. Acids Res. 18:3671, 1990; Syvanen et al., Genoynics 8:684-692, 1990; Prezan et al, Hum. Mutat. 1:159-164, 1992;
Nyren et al., Anal. Biochem. 208:171-175, 1993). These methods differ from Genetic BitTM.
Analysis in that they all rely on the incorporation of labeled deoxyribonucleotides to discriminate between bases at a polymorphic site. In such a format, the signal is proportional to the number of deoxyribonucleotides incorporated, and polymorphisms that occur in runs of the same nucleotide generate signals that are proportional to the length of the run (Syvanen et al Amen. J. Hztm. Genet. 52:46-59, 1993).
[0086] Another method for determining the nucleotide occurrence at a SNP
position is described by Macevicz (U.S. Pat. No. 5,002,867), wherein a nucleic acid sequence is determined via hybridization with multiple mixtures of oligonucleotide probes.
In accordance with such a method, the sequence of a target polynucleotide is determined by permitting the target to sequentially hybridize with sets of probes having an invariant nucleotide at one position, and a variant nucleotides at other positions. The nucleotide sequence is determined by hybridizing the target with a set of probes, then determining the number of sites that at least one member of the set is capable of hybridizing to the target (i.e., the number of matches). This procedure is repeated until each member of a sets of probes has been tested. U.S. Pat. No. 6,294,336 provides a solid phase sequencing method for determining the sequence of nucleic acid molecules (either DNA or RNA) by utilizing a primer that selectively binds a polynucleotide target at a site wherein the SNP is the most 3' nucleotide selectively bound to the target.
[0087] The nucleotide occurrence of a SNP in a sample also can be determined using the SNP-ITTM method (Orchid BioSciences, Inc., Princeton, NJ). lii general, SNP-ITTM is a 3-step primer extension reaction. In the first step a target polynucleotide is isolated from a sample by hybridization to a capture primer, which provides a first level of specificity. In a second step the capture primer is extended from a terminating nucleotide trisphosphate at the target SNP site, which provides a second level of specif city. In a third step, the extended nucleotide trisphosphate can be detected using a variety of known formats, including: direct fluorescence, indirect fluorescence, an indirect colorimetric assay, mass spectrometry, fluorescence polarization, etc. Reactions can be processed in 384 well format in an automated format using a SNPstreamTM instrument (Orchid BioSciences, Inc., Princeton, NJ).
Phase known data can be generated by inputting phase unknown raw data from the SNPstreamTM instrument into the Stephens and Donnelly's PHASE program.
[0088] Melting curve analysis of SNPs (McSNP~ analysis) provides another method for detecting a nucleotide occurrence in an AIM (Alcey ~et al., supra, 2001).
McSNP~ analysis provides the additional advantages that it does not require a step of gel electrophoresis, thus minimizing the time and cost for detecting a SNP, and that it is readily adaptable to high throughput formats, thus allowing examination of one or more panels of AIMs and/or samples in parallel.
[0089] Where the particular nucleotide occurrence of a SNP is such that the nucleotide occurrence results in an amino acid change in an encoded polypeptide, the nucleotide occurrence can be identified indirectly by detecting the particular amino acid in the polypeptide. The method for determining the amino acid will depend, for example, on the structure of the polypeptide or on the position of the amino acid in the polypeptide. Where the polypeptide contains only a single occurrence of an amino acid encoded by the particular SNP, the polypeptide can be examined for the presence or absence of the amino acid. For example, where the amino acid is at or near the amino terminus or the carboxy terminus of the polypeptide, simple sequencing of the terminal amino acids can be performed.
Alternatively, the polypeptide can be treated with one or more enzymes and a peptide fragment containing the amino acid position of interest can be examined, for example, by sequencing the peptide, or by detecting a particular migration of the peptide following electrophoresis. Where the particular amino acid comprises an epitope of the polypeptide~
the specific binding, or absence thereof, of an antibody specific for the epitope can be detected. Other methods for detecting a particular amino acid in a polypeptide or peptide fragment thereof are well lmown and can be selected based, for example, on convenience or availability of equipment such as a mass spectrometer, capillary electrophoresis system, magnetic resonance imaging equipment, and the like.
[0090] In another embodiment, a method of the invention utilizes an antibody, or antigen binding fragment thereof, that specifically binds, for example, to a polypeptide comprising an amino acid encoded by a nucleotide sequence comprising one nucleotide occurrence of a SNP, but not substantially to a polypeptide comprising an different amino acid encoded by the codon comprising the SNP; or that specifically binds, for example, to a polypeptide comprising an amino acid sequence encoded by one form a DIP (e.g., that having the inseuion), but not substantially to that encoded by the alternative form (e.g., that having the deletion). As used herein, the term "specific interaction," or "specifically binds" means that two molecules form a complex that is relatively stable under physiologic conditions. The term is used herein to refer to various interactions, including, for example, the interaction of an antibody that binds a target polynucleotide including the SNP site only if the SNP has a specified, but not an alternative, nucleotide occurrence (e.g, an A, but not a T); or the interaction of an antibody that binds a polypeptide that includes one amino acid that is encoded by a codon that includes a SNP site, but not a polypeptide having an alternative amino acid encoded by the codon comprising the SNP.
[0091] A specific interaction can be characterized by a dissociation constant of at least about 1 x 10'6 M, generally at least about 1 x 10-7 M, usually at least about 1 x 10-$ M, and particularly at least about 1 x 10-9 M or 1 x 10-1° M or greater. A
specific interaction generally is stable under physiological conditions, including, for example, conditions that occur in a living individual such as a human or other vertebrate or invertebrate, as well as conditions that occur in a cell culture such as used for maintaining mammalian cells or cells from another vertebrate organism or an invertebrate organism. Methods for determining whether two molecules interact specifically are well known and include, for example, equilibrium dialysis, surface plasmon resonance, and the like.

[0092] Antibodies useful in a method of the invention include antibodies that specifically bind polynucleotides that encompass an AIM, or that bind polypeptides that include an amino acid encoded by a codon that includes a SNP or that include amino acids due to an insertion at a DIP site. Such antibodies are selected such that they specifically bind a polypeptide that includes a first amino acid encoded by a codon that includes the SNP loci, but do not bind, or bind measurably more wealcly to a polypeptide that includes a second amino acid encoded by a codon that includes a different nucleotide occurrence at the SNP.
(0093] The term "antibody" is used broadly herein to refer to immunoglobulin molecules and antigen binding portions of immunoglobulin molecules that specifically bind an antigen.
As such, antibodies useful in a method of the invention can be polyclonal, monoclonal, multispecific, human, humanized or chimeric antibodies, single chain antibodies, Fab fragments, F(ab') fragments, fragments produced by a Fab expression library, anti-idiotypic (anti-Id) antibodies, and the like, as well as antigen/epitope binding fragments of such antibodies. Antigen binding fragments of antibodies include, but are not limited to, Fab, Fab' and F(ab')2, Fd, single-chain Fv's (scFv), single-chain antibodies, disulfide-linked Fv fragments (sdFv) and fragments comprising either a VL or VH domain. Thus, antigen-binding antibody fragments, including single-chain antibodies, can comprise the variable regions) alone or in combination with the entirety or a portion of the hinge region, CH1, CH2, and/or CH3 domains. The antibodies can be from any animal origin including birds and mammals, or can be expressed recombinantly, for example, in insect or mammalian host cells or in plants.
(0094] There is much that can be learned today through the use of genetic markers in numerous scientific fields. The use of genetic sequences has become routine for forensics and disease research, but the majority of the benefits from the recently completed human genome project still await discovery. Within the genome exist sequences and patterns of sequences that will prove useful for a variety of purposes including increasing crop yields, extending human life spans, minimizing the suffering caused by drugs and enhancing the quality of our lives through better, more effective and specific treatments.
Until now, biomedical research has been conducted on relatively simple terms.
Nevertheless, more one thousand simple Mendelian traits have been mapped by following the transmission of genetic markers in families.
[0095] Many statistical methods are available for studying genetic traits, including traditional family-based linlcage analysis, variance component methods, sib-pair linkage, measured genotype, transmission disequilibrium, genomic control, and structure analysis.
Some of the genes underlying variation in susceptibility to common diseases (e.g., heart disease, obesity, type 2 diabetes, hypertension, and cancer) eventually will be identified using genetic approaches. However, there are a number of complexities in genetic research on common diseases because many of these conditions are multifactorial (i.e., have several sources of variability in rislc) and polygenic (i.e., result due to the actions and interactions among several genes). Additional difficulties in the study of common diseases can derive from the late onset of symptoms and heterogeneity in etiology. Thus, identifying the genes involved in complex diseases remains one of the greatest challenges in the field of human genetics.
[0096] There has been increased interest in association studies as a useful approach to map common disease and drug response genes (Risch and Merikangas, Science 273:1516-1 S 17, 1996; Jorde, Genoyne Res. 10:1435-1444, 2000; Nordborg and Tavare, Tends Genet.
18:83-90, 2002). Until the present disclosure, however, the implication of ancestry for identifying these genes has not been fully appreciated. As such, the methods of the invention provide a previously undescribed platform for the identification of genes associated with disease susceptibility and drug responsiveness, as well as for the development of advanced forensic methods. As such, compositions and methods are provided for inferring an individual's response to commonly used medications, which, remarkably, is a function of individual ancestry; the disclosed markers and methods are, to a differing extent for each drug, useful for the inference of such response. In addition, compositions and methods are provided for inferring individual and/or group ancestral proportions from knowledge of the individual's or group's DNA sequences. Further, compositions and methods are provided for using knowledge of ancestry relevant DNA sequences to identify disease susceptibility and drug response genes through the MALD process. Also, compositions and methods are provided for qualifying and normalizing study groups for more traditional methods of mapping disease genes. Each of these processes requires an accurate knowledge of ancestry, which can be determined using the methods and compositions disclosed herein.
[0097] The populations that will be best suited for linlcage disequilibrium (LD) mapping has prompted much discussion and debate (see Wright et al., Nat. Genet. 23:397-404, 1999;
Eaves et al., Nat. Genet 25:320-323, 2000; Nordborg and Tavare, supra, 2002;
Kaessmann et al., Amef°. J. Hum. Genet. 70:673-685, 2002). The extent of LD is a complex function of a number of genetic and evolutionary factors such as mutation, recombination and gene conversion rates, demographic and selective events, and the age of the mutation itself Some of these factors affect the whole genome, while others only affect particular genome regions.
Additionally, variation in mutation, recombination, and gene conversion rates throughout the genome are expected to create LD differences between genomic regions (see, for example, Taillon-Miller et al., Nat. Gefaet. 25:324-328, 2000).
[0098] Tt has been proposed that small, isolated and inbred populations will have advantages over other populations, due to the lower heterogeneity and the larger extent of linlcage disequilibrium (Wright et al., supf°a, 1999; Nordborg and Tavare, sup~~a, 2002;
Kaessmann et al., sup~~a, 2002). Other populations well suited for mapping are recently admixed populations (e.g., Hispanics and African Americans), which offer the advantage that LD has been created recently due to the admixture process. Because this LD is recent, it can extend over large chromosomal regions. However, it is also extremely important to control for the genetic structure (inter-individual variation in admixture proportions) present in these populations in order to avoid false positives (Parra et al., sup~~a, 1998;
Lautenberger et al., Amer. J. Hum. Genet. 66:969-978, 2000; Pfaff et al., Amen. J. Hum. Genet.
68:198-207, 2001;
Nordborg and Tavare, supoa, 2002, each of which is incorporated herein by reference).
Interest in admixture mapping has increased in recent years (McKeigue et al., Ann. Hum.
Genet. 64:171-186, 2000; Smith et al., J. Invest. De~yzzatol. 111:119-122, 2001; Collins-Schramm et al., Amef°. J. Hum. Genet. 70:737-750, 2002, each of which is incorporated herein by reference). A general description of admixture mapping is provided below, as axe some details about a statistical approach developed for admixture mapping and its application to skin pigmentation as a model phenotype.

[0099] Admixture generates allelic associations between all marker loci where allele frequencies are different between the parental populations (Chalcraborty and Weiss, P~oc.
Natl. Acad. Sci., USA 85:9119-9123, 1988). These associations decay with time in a way that is dependent on the genetic distance between them. Thus, disease (or trait) risk alleles that axe different between the parental populations can be mapped in admixed populations using special panels of genetic marlcers showing high frequency differences between the parental populations. These marlcers, termed AIMS, are characterized by having particular alleles that are mor a common in one group of populations than in other populations. One measure of the informativeness of such maxlcers is the allele frequency differential, delta (8), which is simply the absolute value of the difference of a particular allele between populations (Chalcraborty and Weiss, supra, 1988; Dean et al., sups a, I994).
[0100] In admixed populations, allelic associations were generated recently and, therefore, are more easily detected for a given sample size because they extend over longer distances than in non-admixed populations (up to 10-20 centiMorgans (cM) or more). The statistical basis of this approach was first explored by Chakraborty and Weiss (supra, 1988) and subsequently by Stephens, Briscoe and O'Brien, who named the method "mapping by admixture linlcage disequilibrium" (MALD; Stephens et al., Amer. J. Hum.
Genet.
55:809-824, 1994; Briscoe et al., J. Hey~ed. 85:59-63, 1994). Further, whether one is using a MALD approach or a more traditional LD approach for genetic research, to eliminate associations of the trait with alleles at unlinked loci, it is necessary to control in the analysis for individual ancestry estimated from the marker data. The SNP sequences (markers; AIMS) and methods disclosed herein (BGA test) are a particularly efficient means by which to accomplish this taslc. An Analysis of Covariance (ANCOVA) test has been employed using the estimate of individual admixture as a conditioning variable to control for the effect of individual ancestry in two ways: 1) leaving out the locus under consideration (ANCOVA/IAE minus marker); and 2) using the complete individual ancestry estimate for the conditioning (ANCOVA/IAE). This method is described in detail herein.
[0101] An alternative approach to exploiting admixture has been developed that, while based on earlier work, has little in common with classical LD mapping, and is more analogous to linkage analysis of an experimental cross (McKeigue, Amen. J.
Hum. Genet.
63:241-251, 1998, which is incorporated herein by reference; McKeigue et al., supra, 2000).
For this reason, the term "admixture mapping" has been proposed as more appropriate than "mapping by admixture linkage disequilibrium". Instead of testing for allelic associations, according to the present methods, the underlying variation in ancestry is modeled on chromosomes of mixed descent to extract all the information about linkage that is generated by admixture. The methods and marlcers disclosed are necessary and sufficient to accomplish this process. Advanced statistical methods are utilized to apply this approach in practice, though tl~e underlying principle on which it relies to detect linkage is straightforward.
Suppose, for instance, that a locus accounts for some of the variation in pigmentation between West Africans and Europeans. If individuals of mixed descent are classified according to whether they have 0, 1 or 2 alleles of African ancestry at this locus, then in a comparison of these three groups with other factors held constant, the mean pigmentation level will vary with the proportion of alleles at the locus that are of African ancestry.
Controlling the analysis for parental admixture eliminates association of the trait with ancestry at unlinked loci and ensures that the comparison is made with other factors held constant.
[0102] To infer the ancestry of the alleles at the locus from the marker genotype, the conditional probability of each allelic state is required given the ancestry of the allele (ancestry specific allele frequencies), e.g., West African or European. There is growing evidence that admixture mapping will be an effective means of gene identification, and it has been reported that, in admixed populations, strong allelic association is observed between linked markers spaced at substantial distances (Parra et al., sups~a, 1998;
Parra et al., Amen~. J.
Phys. Auth~opol. 114:18-29, 2001; McKeigue et al., sup~~a, 2000; Lautenberger et al., supra, 2000; Smith et al., sup~~a, 2001; Wilson and Goldstein, Amen. J. Hum. Genet.
67:926-935, 2000; Pfaff et al., sup~~a, 2001). Given the vexy high levels of association observed over long genetic distances, it is expected that phenotypes different between parental populations because of some genetic factor will also show associations with linked AIMS. A
phenotype that is well suited to apply admixture mapping is skin pigmentation.

[0103] Notwithstanding the power of AIMs for disease gene and forensics analysis, no studies have been conducted to elucidate this power. As disclosed herein, 1) SNPs or deletion/insertion polymorplzisms (collectively referred to as AIMS) in the human genome that are of potential use for drug xesponse, disease gene or forensics research were identified;
2) biochemical and genetic test results are provided that demonstrate these AIMS can be useful for disease gene and forensics reseaxch; 3) the usefulness of AIMS
derived from systematic screens of the human genome in actual drug response, disease gene or forensics research is demonstrated; 4) the usefulness of AIMs derived from systematic screens of the human genome to make an inference as to whether an individual is susceptible to acquire a disease, or to not respond to a drug, is demonstrated; 5) the usefulness of AIMS derived from systematic screens of the human genome to make an inference as to whether a crime scene DNA specimen was derived from an individual of, for example, an 80% European, 10% Afxican and 10% Asian heritage or some other ratio/mix is demonstrated; 6) the usefulness of AIMS derived from systematic screens of the human genome to infer the ancestral proportions of an individual from their DNA (e.g., whether the individual is of 80% European, 10% African and 10% Asian heritage, or some other ratio/mix) is demonstrated; and 7) the usefulness of AIMS derived from systematic screens of the human genome to infer the ancestral proportions of a group of individuals from their DNA (for example, whether the group, which can be a population sample, a family, or a clinically defined group of persons, is of 80% Euxopean, 10% African and 10% Asian heritage, or some other ratio/mix) is demonstrated.
[0104] The present results demonstrate that AIMS are useful for the applications described above, and the sequences exemplified herein, as well as additional AIMS
identified using the methods disclosed herein, enable these applications. The AIMS and methods of the invention are useful for the study of human diseases, drug response, and physical traits and, therefore, provide exceptional commercial potential. For example, in this dawning era of personalized drug prescription and disease risk assessment, the markers and methods of the invention provide the tools needed to proceed in this fledgling industry. As exemplified herein, an individual's response to a particular medication was dependent on the degree to which that individual exhibited a certain population structure (i.e., was of certain ancestral heritage) in addition to, but irrespective of, the person's genotype for drug target or xenobiotic metabolism gene sequences. As such, the compositions and methods of the invention provide a means to predict an individual's likelihood to respond to a particular drug.
[0105] For example, in screen of genetic markers associated with patient response to the cholesterol lowering drug, LipitorTM, in terms of low-density lipoprotein (LDL) response, which is an indicator of favorable response, some of the most powerful markers identified for LDL response to LipitorTM were gene types that are not immediately recognized as relevant for drug response, including, for example, TYR, OCA2, TYRP, FDPS, and HMGCR
(see, also, Intl. Publ. No. WO 03/002721 (PCT/LTS02/20847), and Intl. Publ. No. WO

(PCT/LTS02/38345), each of which is incorporated herein by reference). When combined with markers from genes that are biologically relevant for response, they augment the ability to make accurate inferences of response from the DNA. Each of these markers also is an excellent AIM, indicating that the linlcage of the AIMs to drug response is likely a function of ancestral differences in response proclivity (see Example S). As such, ancestral heritage can be predictive of favorable response to LipitorTM. This association has been observed for almost every type of response (n = 54) to almost every type of drug (n = 23) examined, thus confirming that the inference of drug response can be accomplished, at Ieast in part, through the inference of ancestral proportions. As such, it appears that the genes truly relevant for drug response are a function, at least in part, of individual ancestry, and that the gene sequences relevant for drug response are statistically linked with markers that are informative as to ancestry (i.e., AIMs).
[0106] Screening genomes for the true identity of genes associated with a particular trait such as drug responsiveness is extraordinarily expensive and time consuming.
As such, the use of AIMS for making inferences about individual proclivity to drugs provides a significant short-cut for the rapid development of tests that can be used to match patients with those drugs most appropriate for their genetic constitution. Thus, in addition to being useful for the admixture mapping of disease genes, the disclosed methods and exemplified markers provide tools that can direct treatment protocols by clinicians. The identification of AIMs from publicly available human genome data, and the ability to effectively use the AIMs for the development of patient-drug classification sets, admixture screenng panels and forensics tools, was accomplished using the disclosed method, including screening the SNP database (see, for example, world wide web ("www") at URL "nih.ncbi.nlm.gov") for AIMS;
screening the AIMS against a mufti-ancestral panel of DNA samples to verify those that, indeed, are good ATMs; using the disclosed statistical and software methods for using the AIM sequences to malce biologically relevant inferences; and recognizing that an individual's likelihood to respond to a drug or develop a disease can be predicted through a knowledge of their ancestry, which, in turn, is indicated through the individual's AIM sequences.
[0107] Prior to the present disclosure, individual ancestry could be estimated using two independent methods: a Maximum Likelihood approach (Hams et al, Amer. J. Phys.
Anthf°opol. 70(4):433-441, 1986, which is incorporated herein by reference), and a Bayesian method implemented in the STRUCTURE program (Pritchaxd et al., Genetics 155:945-959, 2000, which is incorporated herein by reference). While the Maximum Likelihood method and the Bayesian method provide point estimates of proportional ancestry or admixture, there are several deficits in these methods that are addressed by the disclosed methods. For example, using the disclosed algorithm (see Example 6; see, also, Table 12, containing flow chant exemplifying algorithm), 1) the most likely groups) from which the individual is derived were estimated simultaneously with the estimate of proportional ancestry;
2) multidimensional confidence intervals were computed and projected, thus reducing the complexity fox presentation; 3) an approach to estimate the number of ancestors and their admixture proportions at each level in the past (parental, grandparental, great-grandparental, etc.) was developed; and 4) proportional BGA affiliation within individuals for more than two BGA groups at a time was derived, thus providing, for example, improved and more accurate forensics applications, as well as allowing for the development of classifiers for quantitative or continuously distributed traits (i.e., not dichotomous), the trait values of which are at least in part a function of BGA.
[0108] Independent methods for classification of individuals into population groups have been developed (Shriver et al., sups°a, 1997, Frudalcis et al., supy~a, 2002, each of which is incorporated herein by xeference). The present methods differ from previous classification methods in that they allow the simultaneous estimation of the best group within which a particular individual would fall, as well as the proportional assignment of the individual to multiple parental groups (Example 6; see, also, Table 12). Thus, where previous methods allowed one to malce the statement that a person is much more lilcely to be African-American than European-American, the present approach allow the same statement, and also provides the proportional ancestry of the individual with confidence intervals (CI);
e.g., 2S% (9S% CI

15-3S%); European ancestry; 75% (9S% CI 60-80%) African ancestry; and 0% (95%
CI
0-6%) Native American ancestry. Further, the confidence intervals can be expressed in multidimensional space to provide a clearer representation of the ancestry measured fox the person in question (see below; see, also, Figure 2). Though methods for constructing such a representation were known, the present disclosure is the first to provide fox the representations to be presented with quantifiable confidence.
[0109] There axe clear differences in the patterns of chromosomal segment ancestry (PCSA) among persons with different ancestral histories (see Figure 1). A
series of AIMs across the chromosomes can facilitate the estimation of the most likely parental combinations that lead to the profile of sequences observed in a given person. One example of where estimates of PCSA is important is in the discrimination of persons of Hispanic ancestry from those having primarily European ancestry with some proportion of recent Native American ancestry. Indeed, this is an important determination as the political and legal rights claimed by and provided to these two groups can depend on their ancestry. Hispanic populations such as Mexican-Americans (MA) have approximately 30-40% Native American ancestry, while the balance is European with a minor portion (S% or so) African ancestry. A
person who is one quarter Native American will have 2S% Native American ancestry and, therefore, will overlap with many MA persons in his level of estimated ancestry. It is expected that PCSA
patterns will be significantly different for these two cases and may provide some of the only genetic evidence that would facilitate an accurate definition of the ancestry in such a case.
As disclosed herein, PCSA can be used in ancestry studies.
[0110] An important step in these determinations is the phasing of the AIMS
along chromosomal segments (see Example 2, Figure 8; see, also Example S, Figures 12 to 16).

Phasing ATMs along the chromosome can be accomplished by several methods, including 1) estimation from the genotypes of the individual, 2) molecular haplotyping (e.g., allele specific PCR combined with genotyping), and 3) single sperm analysis (for female subjects the sperm of a male full sibling would provide the same profile). In addition, the disclosed methods allow simultaneous consideration of the two sex chromosomes (X and Y) and the mtDNA for ancestral inferences. AIMS are found on each of these sources, and can be informative for many of the questions regarding the ancestral proportions of a person and the populations) from which a particular person is derived. For example, Hispanic/Latino populations have very high (65-100%) frequencies of Native American mtDNA
haplogroups, while showing only a minority contribution from Native American populations in autosomal marlcers. Thus, for example, a person with reputed Native American ancestry on her father's side, with a non-Native American mtDNA haplogroup, is more likely not Hispanic than partially Native American as she may suspect, than were she to have a Native American mtDNA haplogroup.
[0111] Linlcage disequilibrium (LD) is increasingly being used as a mapping tool for both fine-scale determination of gene position and for the initial localization of disease genes in special populations. Allelic associations axe significantly non-random and correlated with physical distance within small (< 60 kb) genomic regions (see Jorde, Amey~. J.
Hum. Genet.
66:979-988, 1995; Jorde, Genome Res. 10:1435-1444, 2000, for review), possibly reflecting an underlying "block structure" that characterizes many genomic regions (Reich et al., 2001;
Daly et al., 2001). Thus, if disease alleles in a population share a recent common origin, nearby genetic markers with the strongest associations will be closest to the disease-causing locus. This approach has been important in the positional cloning of several simple Mendelian diseases, including the cystic fibrosis gene, the Huntington's disease gene, and the diastrophic dysplasia gene.
[0112] Tn addition to applications in fme-mapping or positional cloning, LD
can be used for initial disease gene mapping in homogeneous populations that have undergone recent increases in size or are genetically inbred. In such populations, disease alleles were probably present in a small number of founders, and recombination has had limited opportunity to randomize associations between these alleles and alleles at linked marker loci. An analysis of allelic associations between affected and unaffected individuals from these populations can thus facilitate the localization of the disease locus. A number of Mendelian diseases have been mapped using this approach: several diseases in the Finish population, Hischsprung's disease in Mennonites, benign recurrent intrahepatic cholestasis in an isolated Dutch fishing comunmlity, familial persistent hyperinsulinemic hypoglycemia of infancy in a consanguineous group of Saudi Arabian families, and Bardet-Biedl syndrome in Bedouins.
[0113] There has been much debate as to which populations will be best suited for LD
mapping of complex polygenic diseases (see, e.g., Wright et al., supy~a, 1999;
Eaves et al., sups°a, 2000; Nordborg and Tavare, supra, 2002; Kaessmann et al., supra, 2002). The extent of LD is a complex function of a number of genetic and evolutionary factors such as mutation, recombination and gene conversion rates, demographic and selective events, and the age of the mutation itself. Some of these factors affect the whole genome while others only affect particular genome regions. Additionally, variation of mutation, recombination and gene conversion rates throughout the genome is expected to create LD
differences between genome regions.
[0114] With regard to the populations to use in disease discovery efforts driven by LD-based methods, it has been proposed that small, isolated and inbred populations will have advantages over other populations, due to the lower heterogeneity and the larger extent of linkage disequilibrium (see, e.g., Wright et al., supoa, 1999; Nordborg and Tavare, supra, 2002; Kaessmann et al., supy~a, 2002). On the other side, admixed populations such as Hispanics and African Americans offer the advantage that linkage disequilibrium has been created recently due to the admixture process, and it can extend over large chromosomal regions, although it is extremely important to control for the genetic structure present in these populations in order to avoid false positives (Parra et al., supra, 199;
Lautenberger et al., supra, 2000; Pfaff et al., sups°a, 2001; Nordborg and Tavare, supra, 2002). Despite the increased research focus in LD based methods, however, many issues regarding LD in human populations remain largely unexplored. Currently the NHGRI is organizing a systematic project to help develop informational tools for gene identification studies by identifying the common haplotypes in several populations. This "Haplotype Map Project" (HMP) will Iikely be a large-scale multicenter effort focused on finding common haplotypes in general population samples. The HMP will likely prove to be an important data resource for identifying AIMs as disclosed herein because several populations will be investigated for haplotype bloclc structure, thus providing additional candidate AIMs and a basic plan for the fine scale LD structure in some of the parental populations.
[0115] Admixture mapping as disclosed herein is complementary to, but distinct from, the HMP. First, the primary focus of the HMP is to understand the fine scale structure of individual genomic regions throughout the genome, whereas the present methods allows an understanding of the LD that results specifically from admixture. The level of LD from admixture is on the order of millions of bases (Mb; megabases) and tens of Mb, while the HMP is focused on the level of 10's to I00's of Icilobases (Icb), and genomic and population features that affect the results from one project may not be noted in the other. Second, admixture mapping require accurate parental allele frequency estimates. As such, a large number of different African, Native American, European, and Asian populations have been typed (see Table 6, below), while the HMP will likely focus on one or two samples of the major population groups.
[0116] Third, Iarge samples (n = 500) of African-Americans and Hispanics have been typed, thus providing sufficient statistical power to test the coverage of the admixture map and to compare analytical methods. In addition, several representative populations from different regions of the country were typed so that geographical variation in ancestral proportions and admixture dynamics can be examined. Although some admixed populations will likely be included in the HMP, the numbers of individuals and numbers of different population samples being discussed are fewer than those as disclosed herein and, therefore, will not allow the same comparisons. For example, having a sample of 10 for each of 4 ancestral groups is not adequate for the identification of sequences present preferentially in one or some of those groups; as disclosed herein, at least 50 individuals were tested for each of several tens of ancestral groups (not just four) in order to comprehensively identify these markers.

[0117] Fourth, the focus of current population variation efforts (e.g., the SNP Consortium allele frequency project) and, very likely, the HMP has been on East Asian, African, and European samples to the exclusion of Native American populations for a number of complex reasons. The exclusion of these populations, however, results in a deficit in an understanding of the genetics of the fastest growing group of US resident populations, i.e., Hispanics, who have a siguf cant level of Native American ancestry (20% to 40%). With the markers and methods disclosed herein, the disease genetics of Hispanic populations can be examined.
Similarly, several diverse Native American populations may represent important parental populations for the numerous distinct groups often grouped together as Hispanic.
[0118] The population-based association methods disclosed herein provide several advantages over traditional linkage studies. Localizing disease genes by traditional genetic linkage methods relies on the use of related persons, either extended multigenerational families or pairs of related individuals. These approaches are effective and very powerful when investigating diseases caused by single genes. However, polygenic and multifactorial diseases lilce Type 2 diabetes, hypertension, and prostate cancer result from the interaction of several genes and multiple environmental influences, and are more difficult to study using traditional methods. The identification of genes contributing to susceptibility to common disease is complicated by heterogeneity. The source of the genetic heterogeneity determines which mapping methods are most likely to work for gene identification. Two primary types of genetic heterogeneity are locus heterogeneity, wherein more than one locus is affecting a genetic trait, and allelic heterogeneity, wherein within a particular causative locus there are multiple alleles that are important in altering the phenotype. Traditional linkage analysis using extended families is generally insensitive to allelic heterogeneity, but can be adversely affected by locus heterogeneity. Alternatively, LD based methods are generally adversely affected by allele heterogeneity, but Iess affected by Iocus heterogeneity.
[0119] Provided there is little allelic heterogeneity, association-based approaches like measured genotype and transmission disequilibrium test (TDT) may be more sensitive than family-based LOD score or sib-pair methods. Risch and Merikangas (supra, 1996) compared the number of individuals needed for sib-pair studies and TDT studies, and showed that the number of individuals needed to detect linkage is much smaller for TDT than for sib-pair studies. This is especially true when the disease locus has a small effect.
For example, for a locus with risk ratio of 2.0 and a gene frequency of 50%, 2500 sib-pairs or 340 case/parents for TDT would be required. There are some examples in which the demonstration of association using Haplotype Relative Risk (HRR) or caselcontrol design, or linkage with TDT, has preceded the demonstration of linkage with sib-pairs. A classic example is the association between the insulin gene and IDDM, which was demonstrated in cases and controls, then confirmed using TDT, but often not observed in sib-pair linkage studies (reviewed in Spielman et aL, Amen. J. Hum. Gefzet. 28:317-331, 1993). Yaouanq et al.
(Sciev~ce, 1997) reported very significant (p < 10-9) evidence for linkage between the HLA
and multiple sclerosis using TDT in a series of 157 French families (99 simplex and 58 multiplex). When the 58 multiplex families were analyzed alone, p values of 0.0001 and 0.03 resulted for TDT and sib-pair methods, respectively.
[0120] Although association studies based on candidate genes have relatively higher power to detect disease genes than linkage analysis in families (Risch and Merikangas, sups°a, 1996), thoroughly testing all the genes in a genome with over 40,000 genes is currently not practical. The Haplotype Map Project may succeed in creating the informational resources necessary to perform gene identification based on linkage disequilibrium.
However, even if the block structure models of the human genome can be explained by four haplotypes in each gene, the minimum number of SNPs and DIPS would then be 80,000 and the actual number likely higher. Although genotyping technologies are advancing rapidly, typing this number of markers in a large number of research subjects is not yet practical.
Additionally, there are some important assumptions that are implicit in plans to identify genes using LD in large populations. One important difficulty using linl~age disequilibrium in genome-wide screening is that LD decays exponentially with the recombination fraction between the marker and the disease locus and with the age of the disease-causing mutations. For older mutations that predispose to diseases, LD becomes very weak even between the disease allele and alleles at relatively closely spaced marlcer loci.

[0121] LD mapping has been useful in mapping of rare genetic diseases such as cystic fibrosis and diseases in special populations like the Finns and Bedouins, populations that have been subject to significant population bottlenecks, inbreeding, or founder effects. In these situations, LD exists because the variant allele is relatively young, as in the case of cystic fibrosis, or the population has reduced genetic variability, which elevates the LD
throughout the genome. A leading model for the genetics of common disease stipulates predisposing alleles at a number of loci which, when present in particular combinations, increase an individuals xislc (Greenberg, Amen. J. Hum. Genet. 52:135-143, 1993; Lander and Schorlc, Science 265:2037-2048, 1994; Risch and Merilcangas, supra, 1996). If the disease is common, then for this model, the predisposing alleles also are expected to be at relatively high frequencies. However, assuming the neutral model, the frequency of an allele in a population is on average related to the age of the allele such that more frequent alleles are older than rare alleles. This fact poses a problem for the application of LD-based methods to identify common disease genes in populations that axe not isolated or inbred since, in homogeneous populations, the LD is inversely related to the age of the allele and risk alleles for common disease are expected on average to be relatively old.
[0122] The application of the compositions and methods of the present invention for admixture mapping allows for a more precise and reliable mapping of complex traits.
Admixture mapping takes advantage of the LD created when previously isolated populations admix, and can circumvent these problems in mapping complex traits. It was first recognized that admixed populations could be useful in determining genetic linlcage by Chakraborty and Weiss (supoa, 1988). When genetically divergent populations hybridize, non-random allelic associations result among loci that have significant allele frequency differentials, even among unlinked loci. This LD quickly decays when the genetic loci in question are not located close together on the same chromosome.
[0123] LD decays as a function of the recombination rate (8) between the two markers and the riumber of generations (n) since their hybridization, and can be represented as D" _ (1-0)"
Do, where D" is the linkage disequilibrium n generations after hybridization and Do is the initial linkage disequilibrium (Chakraborty and Weiss, supra, 1988). Given this exponential relationship between the decrease in LD and genetic distance, it is possible to discriminate between LD in an admixed population (if the time since admixture is short) that remains high because markers are close together, genetically linked, and background linkage disequilibrium among unlinked loci. For example, after 10 generations, the linkage disequilibrium at unlinlced loci is reduced to 0.1 % of the initial level, while at loci 10 cM and 1 cM apart, the disequilibrium due to true linkage will still be 34.9% and 90.4%, respectively, of the initial level. The critical parameters for effective detection of linkage in an admixed population identified are the frequency differential (8) between the parental populations and the number of generations since hybridization. Linkage by association analysis in admixed populations worlced efficiently if 8 was large (not less than 0.2) and the number of generations since admixture small (on the order of 10 generations;
Chalcraborty and Weiss, sup~~a, 1988).
[0124] Stephens et al. (sup~~a, 1994) and Briscoe et al. (supra, 1994) studied this approach using computer simulations (MALD) and detailed practical considerations for study design.
Using simple models of the type of admixture that has occurred in the Americas; they suggested that using sample sizes of 200-300 patients, typed for 200-300 evenly spaced markers, each having ~ > 0.3, one would have > 95% chance of locating the causative gene.
A consistent result from the several models studied was a primary dependence of the power of MALD on the allele frequency differentials of the markers used. Tf 8 is small, the initial LD will be small and difficult to distinguish from the background noise.
[0125] Stephens et al. (sups°a, 1994) suggested using loci where 8 >
0.4 between the admixed parental populations for effective admixture mapping. They also demonstrate that admixture mapping is most effective in populations that hybridized between 4 and 20 generations ago, and that incremental admixture (the slow introgression of one population into another; also known as the continuous gene flow model) affects the power of admixture mapping, but not critically, provided there has been no new introgression from the parental populations in the past three generations. The disclosed admixture mapping technique can identify the location of disease susceptibility genes by the analysis of admixed populations composed of parental populations where there is a large difference in the frequency of the susceptible genotype. As such, applications of admixture mapping include the study of Type 2 diabetes susceptibility in Pacific Island populations, hypertension obesity, and prostate cancer in African Americans, and Type 2 diabetes, obesity and gallbladder disease in Hispanic populations.
[0126] McKeigue developed an approach to exploit admixture in mapping genes that builds on earlier work (McKeigue, supra, 1997; McKeigue, supra, 1998, McKeigue et al., supra, 2000). Although the approach is powered by the LD that is generated by admixture, it is more analogous to linkage analysis of an experimental cross. For this reason, the term "admixture mapping" was proposed. Instead of testing for allelic associations, one can model the underlying variation in ancestry on chromosomes of mixed descent to extract all the information about linkage that is generated by admixture.
[0127] As discussed above, advanced statistical methods are required to apply this approach in practice. Conditioning on parental admixture eliminates association of the trait with ancestry at unlinked loci and ensures that the comparison is made with other factors held constant. In non-statistical language, a comparison is made in each individual of the proportion of alleles at the marker locus that are of a particular descent with the expected proportion given the admixture of that individual's parents. One simple way to do this is to use the Analysis of Covariance (ANCOVA) test (see Tables 2 and 3), though this simpler approach does not use all of the information available. As such, Bayesian methods also have been used (see Tables 2 and 3).
[OI28] To infer the ancestry of the alleles at the locus from the marker genotype, the ancestry-specific allele frequencies are required; i.e., the conditional probability of each allelic state given the ancestry of the allele (West African or European, in this example). The total population of alleles at any locus in the admixed population can be considered to be made up of two subpopulations - alleles of African ancestry and alleles of European ancestry.
As long as the ancestry-specific allele frequencies are correctly specified for the admixed population under study, Bayes' theorem can be applied to invert these conditional probabilities and calculate the posterior distribution of ancestry at the locus (0, 1 or 2 alleles of African ancestry) for each individual under study. If the information conveyed by typing a single marker is not sufficient to assign the ancestry of each allele at the marker locus to one of the two founding populations, marlcers can be combined in a multipoint analysis to estimate ancestry at adjacent loci.
[0129] Simulation studies showed that, with enough markers, a high proportion of information about ancestry at each locus can be extracted even though no single marker is fully informative for ancestry (McKeigue, supra, 1998). Based on these simulations, panels of markers with Fst >0.4 at an average spacing of 2-3 cM for a total of 1,000 AIMS can be constructed as disclosed herein. It should be recognized that the panel of 1,000 AIMS for a particular population (e.g., an African-American group that is primarily West African and European) will often overlap with panels for other groups. In other words, it is often the case that AIMS selected for one level of distinction (e.g., African/European) are also informative for other distinctions (e.g., Native American/European). Table 1 lists an initially identified panel that includes of 32 ATMs (SEQ ID NQS:332 to 363; see, also, Example 1).
Using a cutoff of d > 0.3, only four of these markers are restricted in informativeness to one of the three comparisons (African/European; African/Native American; Native American/European); the rest are informative for two of the comparisons, and one marker is informative for all three comparisons. In a further study, a panel of 71 AIMS
was identified (SEQ ID NOS:1 to 71; Table 6) that are informative as to IndoEuropean, sub-Saharan African, Native American, and East Indian (see Example 2).
[0130] There is growing evidence that admixture mapping will be an effective means of gene identification. At least three independent groups have reported strong admixture linkage disequilibrium (ALD) between linked marlcers spaced at substantial distances (see, e.g., Parra et aL, supra, 1998 and 2001; Pfaff et aL, supra, 2001; McKeigue et al., sups°a, 2000). Given the very high levels of association that have been observed over long genetic distances, it is expected that phenotypes dramatically different between parental populations because of some genetic difference will also show associations with linked AIMS. However, as promising as that MALD approach appears, until the present disclosure, no systematic screen has been reported identifying SNP based versions of the AIMS required.
McKeigue and others have identified panels of STR AIMs for use with this approach, but the use of STRs for this purpose is problematic because of the allelic complexity of STRs and the massive databases required in order to accurately estimate allele frequencies.
Even small errors or faulty assumptions on the frequencies of unobserved alleles can amplify to cripple the statistical power of a study.
[0131] Heterogeneity within the parental populations can have a confounding effect on admixture mapping studies. In the case of African-American populations, the process of admixture that tools place in the New World involved a heterogeneous group of populations mainly from West-Central Africa and Europe, as well as some Native American populations.
Regarding the European genetic contribution, the most important source populations came from Great Britain, Ireland, Germany and Italy. In spite of the diverse geographical areas of origin of the parental European populations, it is important to indicate the relative homogeneity of European populations from the genetic point of view (see for example Cavalli-Sforza et al., supra, 1994).
[0132] With respect to the African contribution, it is well known that the African continent contains a tremendous amount of genetic diversity. However, only a subset of the African genetic diversity contributed to the formation of African-American populations. The majority of enslaved Africans came from West-Central Africa, approximately from Senegal in the North, to Angola in the South (Cumin, In The Atlahtic Slave Trade;
Madison, University of Wisconsin Press 1969); other areas of Africa were not affected by the slave trade. Of the four main linguistic families present in Africa, Niger-Congo Kordofanian, Nilo-Saharan, Afro-Asiatic and Khoisan (Greenberg, supf°a, 1963), the majority of enslaved Africans forcefully brought to the New World were members of the Niger-Congo family.
This widespread family encompasses West African languages (spoken by peoples from Senegal to Nigeria) and Bantu languages (dominant in Central and Southern Africa). The Bantu languages were dispersed throughout Africa by a "recent" expansion that took place about 3,000 years ago, and probably originated in West Africa (Nigeria and Cameroon; see Excoffier et al., Yea~bookPl~ys. Authropol. 30:151-194, 1987 and Cavalli-Sforza et al., supra, 1994). This recent origin is reflected in the linguistic and genetic homogeneity of the Bantu (Excoffier et al., supra, 1987, Weber et al., supT a, 2000). Thus, the available Sb historical, linguistic and genetic evidence indicate that only a subset of the diversity found in sub-Saharan Africa has contributed to the African-American gene pool, and that potential problems of heterogeneity are much less than if the diversity of the whole continent of Africa were represented in contemporary African-American populations. Unfortunately, the extent of the heterogeneity present in West and Central Africa remains largely unknown due to the lack of available information for the populations of this area.
[0133] Since the extent of heterogeneity within European populations and within West and Central Africa remains largely unknown, potential effects of heterogeneity need to be addressed, particularly when considering an admixture mapping approach. There are two levels at which heterogeneity can affect an admixture mapping effort. First, heterogeneity can lead to erroneous estimates of the parental frequencies for the markers used in the map, thus biasing the estimate of admixture. Given that the goal of admixture mapping is to infer linkage conditioning on parental admixture, it is important to avoid misspecification of the ancestry-specific allele frequencies, because this could affect the final outcome of the analysis. Second, heterogeneity can affect the number of loci for the phenotype being studied.
[0134] The effect of heterogeneity in biasing the estimates of the genetic contributions to an admixed population can be reduced by selecting markers showing homogeneity within the main parental populations (Europeans and Africans). In this way, the problem of contribution of different geographical axeas to the parental populations is minimized, reducing the bias in admixture estimates. This strategy was implemented in previous admixture studies (Parra et al., sup~~a, 199, 2001; Pfaff et al., sup~~a, 2001), wherein potentially informative markers in different European and African populations were systematically analyzed. As am example, currently, to test for heterogeneity within Africa, each potentially informative marlcer was genotyped in samples from five African populations, two from Nigeria, two from Sierra Leone and one from Central African republic, and the markers showing significant heterogeneity were excluded from the analysis (see below). All of these samples came from areas that were affected by the slave trade. If desired, a sample from Angola, which is a region that was the source of around 40% of enslaved Africans, can be incorporated, thus providing another sample of African parental populations. In addition to this strategy, it is important to note that there are statistical methods to test for misspecification of parental frequencies (see, e.g., McI~eigue et al., supra, 2000).
[0135] With respect to the potential problem of heterogeneity in the phenotypes being studied, it is expected that heterogeneity due to the presence of multiple genes (locus heterogeneity) affecting a phenotype will reduce the power of admixture mapping to detect significant genotypic effects, as it does with any other mapping method.
Heterogeneity also can be due to multiple functional alleles within a particular gene (allelic heterogeneity). One example is MC1R, Where approximately six relatively common variants Iead to red hair, freclcles, and pale skin among Native Europeans and their descendant populations. Within Europeans these variants are on different haplotype backgrounds, thus decreasing the power to detect an effect of the MCIR gene in association studies relative to the case where a single mutation had occurred and risen to high frequencies. However, in an admixed population (e.g., European/African), these variants will all be in allelic association with marlcers informative for ancestry (e.g., the MC1R marker, see Table 1) and, since they all have the effect of lightening the skin, their information will be compounded making the identification of MC 1 R by admixture mapping no different with six functional variants than were there only one functional variant unique to Europeans. So long as the effects of functional variants within a particular parental population are in the same direction (for example, in lowering the risk of disease), allelic heterogeneity will not be a serious problem in admixture mapping.
[0136] The majority (80-90%) of genetic variation among human individuals is inter-individual; only 10-20% of the variation is due to population differences (e.g., Nei, sups°a, 1987; Cavalli-Sforza et al., supra, 1994, Delca et al., supra, 1995). Most populations share alleles and those alleles that are most frequent in one population generally are also frequent in others. There are very few classical (blood group, serum protein, and immunological) or DNA genetic markers which are either population-specific or have large frequency differentials among geographically and ethnically defined populations (Roychodhury and Nei, supra, 1988; Cavalli-Sforza et al., supra, 1994). Despite this apparent lack of unique genetic markers, there are marked physical and physiological differences among human populations that presumably reflect long-term adaptation to unique ecological conditions, random genetic drift, and sex selection. In contemporary populations, these differences are evident both in morphological differences between ethnic groups and in differences in susceptibility and resistance to disease.
[0137] The most useful unique alleles for admixture and mapping studies are those that also have large differences in allele frequency among populations (Reed, supra, 1973;
Chalcraborty et al., sup~~a, 1992; Stephens et al., supra, 1994). The fact that they are totally absent from all other populations does simplify some of the statistical computations, and can facilitate more conf dent parental allele frequency estimates, but is not the primary reason for their utility. The designation population-specific alleles (PSAs) was initially used to describe genetic markers with large allele frequency differentials between populations (Shriven et al., supra, 1997; Parra et al., supra, I998), but these markers are now referred to by the more correct and descriptive term, Ancestry Informative Markers (AIMS). For a biallelic marker, the frequency differential (~) is equal to pX - py, which is equal to qy - qX, where pX and py are the frequencies of one allele in populations X and Y and qX and qy are the frequencies of the other. Median ~ levels among major ethnic groups range between 15% and 20%, and the vast majority (> 95%) of arbitrarily identified biallelic genetic markers have 8 <
50% (Dean et al., supy~a, 1994, which is incorporated herein by reference). Statistical estimates of power in an admixture mapping study based on using markers with an Fst > 0.4 were previously presented (McI~eigue et al., sup~~a, 2000). With 1,000 such markers evenly spaced across the genome, it was demonstrated that it was possible to have a statistical power of 80% to identify a disease gene that explains a 2 fold relative risk between the parental populations.
[0138] AIMS, and their use according to a method of the invention are demonstrated in Exaanples 1 to 6 (below). In addition, allele frequency data for markers informative for admixture mapping in African-American and Hispanic populations have been reported (Smith et al., sups°a, 2001; Collins-Schramm et al., supfa, 2002). In order to apply the methods of the invention to an analysis of disease predisposition or drug responsiveness, estimation of admixture proportions and admixture dynamic is important.
Controlling for genetic structure in admixed populations requires knowledge of the ancestral proportions and the genetic structure of these populations. Reliable estimates of admixture proportions can allow the informed identification of populations to consider. Since the admixture LD that is created during hybridization is dependent on the level of admixture, sampling should focus generally on those areas of the country where there has been more admixture.
[0139] In addition to knowing the ancestral proportions, it is important to understand the level of population structure present in the populations under consideration.
A homogeneous population is one in which there is no assortive mating, a panmictic population in which families are formed more or less by random combination and without regard to DNA
genotypes. In most large cosmopolitan populations, homogeneity is expected and found. If, however, there exists stratification within the population such that individuals do not mate at random, the population will not be homogeneous. Admixture is one of the possible mechanisms introducing genetic structure in a population, and taking into account this genetic structure facilitates admixture mapping.
[0140] The effect of genetic structure is considered at two levels. First, parental populations are evaluated to determine whether they show heterogeneity in the allele frequencies of the selected AIMs; heterogeneity can affect the estimate of admixture proportions, as discussed above. Several methods that can detect the presence of genetic structure. These methods can be grouped in two main categories, termed genomic control (GC) methods (Devlin and Roeder, supra, 1999), and structured association (SA) methods (Pritchard and Donelly, supra, 2001). Both methods require genotyping of a panel of unlinked markers to estimate and correct for the effect of genetic structure, which, as discussed above, rnay have been due to sampling effects, or due to real demographic strata in the sampled population. The SA method (Pritchard et al., supra, 2000;
Pritchard and Donelly, sups°a, 2001) was used to test for genetic structure in the parental populations. This method is based on using the genotypic information provided by the unlinked markers to infer population structure, and has been implemented in a software program available from Jonathan Pritchard. In addition, to test for the presence of stz~ucture, the program estimates individual ancestry proportions, and, for the present studies, this Bayesian method was used to complement the Maximum Likelihood Estimate method. These two methods produce estimates of individual ancestry that are highly correlated.
[0141] The second source of genetic structure in admixed populations is due to the admixture process itself, in which newly created linkage disequilibrium is introduced in the admixed population. AIMS, such as those exemplified herein, are particularly sensitive indicators of population structure that is related to ancestral proportions.
To evaluate the presence of population structure, samples are tested for the non-random association of alleles both within a locus (Hardy-Weinberg disequilibrium) and among loci (gametic disequilibrium), and the distribution of individual ancestry estimates also is examined (see, Pfaff et al., sups°a, 2001; Parra et al., supra, 2001).
[0142] The history of African Americans can be traced back to 1619, when the first Africans arrived at the British colonies (Jamestown), although as early as 1526 the presence of African slaves was reported in Spanish expeditions to what would become the United States (South Carolina, Georgia, Floxida and New Mexico). Institutional slavery began very soon after, but it was not until the beginning of the 18th century that the importation of slaves reached increased rates, in parallel with the demand for workers to cultivate the tobacco, indigo, and rice plantations in the southern colonies; peaks occurred in the decade from 1790-1800 and the first years of the 19th century. In 1808, slave trade became illegal but continued at a low rate for several more decades. Different estimates have been offered on the total number of slaves brought into the United States with generally accepted numbers ranging between 380,000 and 570,000.
[0143] Although it is difficult to precisely determine the ethnic origins of the African slaves, information from shipping lists has provided an approximate picture of their geographic provenance. The slave trade affected a very wide area of Western and Western-Central Africa, mainly the coastline between the present day countries of Senegal in the North and Angola in the South. The most important regions were Senegambia (Gambia and Senegal), Sierra Leone (Guinea and Sierra Leone), Windward Coast (Ivory Coast and Liberia), Gold Coast (Ghana), Bight of Benin (From the Volta river to the Benin river), Bight of Biafra (East of Benin river to Gabon), and Angola (Southwest Africa, including part of Gabon, Congo and Angola). Cumin (sups°a, 1969) has offered, based on data on the English trade of the 18th century (the peak of the Atlantic slave trade), estimates of the proportional contribution by areas, showing that Angola and Bight of Biafra were the regions contributing the highest numbers of slaves imported into the North American maiuand (around 25% each). However, there were significant differences in ethnic origin depending on the port of entry in the United States, and the figures for the Colonies of Virginia and South Carolina differed considerably.
[0144] The history of African Americans has bean marked not only by the forced migration from Africa, but also by the admixture with the other ethnic groups they met when they arrived in North America, including with Europeans and Native Americans.
However, few historical records address the issue of admixture. Additionally, there have been important factors that, in the time since the abolition of slavery until the present, have configured the present African-American population. Of special interest is the pattern of migration of African Americans within the United States over the past 150 years. In this sense, the redistribution of African Americans in the Southern States during the 19th century, and the Crreat Migration from the rural South to the urban areas in the North beginning after World War I are of particular relevance, and have had an enormous impact in defining the present distribution of the African-American population in the US (Johnson and Campbell, In Black Migration ih Ame~icasz: A Social Demographic History; Duke University Press, Durham NC 1981).
[0145] With respect to Hispanics, the term "Hispanic" was coined mainly for governmental demographic purposes, and is generally employed to identify persons of Latin American origin or descent, living in the United States. Although this definition lumps together people with very different historical, cultural and linguistic backgrounds, this classification has been widely used. Even though Central America, the Caribbean, and South America have been for centuries under the domination of the Iberian imperial powers (Spain and Portugal), they have had quite different regional histories, both before and after the Colonial period. Populations from four continents, North and South America, Europe, and Africa, have contributed to the formation of contemporary Hispanic populations. The anthropological background of the main three Hispanic groups currently living in the United States - Mexican Americans, Puerto Ricans and Cuban Americans, which together makeup more than 80% of the total US Hispanic population - is considered here.
[0146] Mexican Americans show the highest Amerindian contribution of the three aforementioned groups. Soon after the Spanish conquest of Mexico, at the beginning of the 16th century, intermixture of the Spanish men with Amerindian women resulted in an increasingly important mixed population (Mestizos), and this racial mixing continued through the three centuries of Spanish domination in "New Spain", configuring both biologically and culturally the Mexican population. The majority of estimates have indicated an Amerindian component in Mexican Americans ranging between 30% and 40% (Harris et al., supra, 1986;
Long et al., 1991; Harris et al., Diabetes Ca~~e 14:618-627, 1991; Merriwether et al., Anzeo. J.
Phys. Av~throp. 102: I53-159, 1997). It is interesting to point out, as well, that some studies have shown differences in the amount of Amerindian ancestry depending on socioeconomic status (Chakraborty et al., Geszet. Epidefniol. 3:435-454, 1986; Mitchell et al., Ethnicity ahd Disease 3:22-31, 1992). There was also a substantial African presence in the Mexico territory during the Spanish rule. Curtin (supj~a, I 969) has estimated the total number of Africans imported into Mexico during the entire period of Slave trade to be around 200,000.
Their contribution to the Mexican gene pool, however, has been estimated to be much lower than the European and Amerindian contribution, ranging from zero to 10% (see, e.g., Harris et aL, supra, 199I).
[0147] In the Caribbean Colonies (Cuba and Puerto Rico), the situation was very different from the Mainland. The Native American population was far smaller, and was decimated by slavery and disease very soon after the first contact with the Europeans.
Nevertheless, the rate of admixture during the initial phases of the colonization was high enough to result in an appreciable genetic contribution (about 18%) from the Arawal~s and Caribs, the original inhabitants of the Hispanic Caribbean (Harris et al., supra, 1991). Another distinctive feature of this region is a significant African influence, which is also reflected in many aspects of the present societies of countries like Cuba, Puerto Rico, and the Dominican Republic. African slaves were imported to work in the sugar plantations in large numbers, even outnumbering the population of European origin (Kanellos and Perez, In Chronology of Hispanic American histoT y: from p~~e-Columbian tiyfZes to the p~~esent; New Yorlc, Gale Research 1995).
Accordingly, the percentage of African genetic contribution in contemporary Cubans (20%) and Puerto Ricans (37%) is significantly higher than in other Hispanic populations (Hams et al., sups~a, 1991).
[0148] It is clear that race is a complex concept and, in general usage, reflects both a cultural and biological feature of a person or group of people. Given the fact that physical differences between populations are often accompanied by cultural differences, it has been difficult to separate these two elements. There has been a movement in several fields of science to oversimplify the issue declaring that race is merely a social construct. While this often can be true, depending on what aspect of variation between people is being considered, it can be false for many particular instances of differences between the populations of the world. One clear example of a biological difference is skin color. Culture or environment has almost no effect on the level of pigmentation in a person's skin. Yet there are dramatic differences across populations. Pigmentation is, of course, only skin deep and is quite simple in light of the complex environments in which we live and how these affect individual and group quality of life.
[0149] The human species is relatively young and, as a species, most likely originated in east Africa 100,000 years ago, and diverged as groups to settle the globe (Cavalli-Sforza and Cavalli-Sforza, In The Great Human Diaspo~~as. The Histo~~y of Dive~~sity and Evolution (Perseus Boolcs, Cambridge MA 1995). During these migrations, and in the time since, there has been some degree of independent evolution of the populations that settled the various continents of the world. The simplest evidence of this evolution is seen in the differences in allele frequencies at genetic marlcers. Generally, alleles that are found in one population are also found in all populations, and the alleles that are the most common in one population also are common in others. These similarities between populations highlight the recent common origin of all populations. However, there are examples of genetic markers that are different between populations and, as disclosed herein, these markers, AIMS, can be used to estimate the Ancestral origins of a person or population.

[0150] The present invention provides methods of estimating proportional ancestry of at least two ancestral groups of a test individual and, in particular, provides a confidence level with respect to the proportional ancestry. A method ~of the invention can be performed by contacting a sample, which includes nucleic acid molecules of the test individual, with hybridizing oligonucleotides that can detect nucleotide occurrences of SNPs of a panel of at least about ten AIMS that are indicative of BGA for each ancestral group examined, wherein the contacting is ~mder conditions suitable for detecting the nucleotide occurrences of the AIMS of the test individual by the hybridizing oligonucleotides; and identifying, with a predetermined level of confidence, a population structure that correlates with the nucleotide occurrences of the AIMs of each of the ancestral groups examined, wherein the population structure is indicative of proportional ancestry.
[0151] The term "biogeographical ancestry" or "BGA" is used herein to describe the biological or genetic component of race. BGA is a simple and objective description of the ancestral origins of a person, in terms of the major population groups (e.g., Native American, East Asian, Indo-European, and sub-Saharan African). BGA estimates can represent the mixed nature of many people and populations today. In many countries, including the United States, there has been extensive mixing among populations that initially had been separate.
The term "admixture" is used herein to refer to such population mixing. In this respect, BGA
estimates can be understood as individual admixture proportions, which take the form of a series of percentages that add to 100%. For example, a person can have 75%
Indo-European, 15% African, and 10% Native American ancestry, or can have 100% Indo-European ancestry, or the lilce.
[0152] The proportional ancestry estimated according to a method of the invention can be a proportion of any ancestral group, including, for example, a proportion of sub-Saharan African, Native American, IndoEuropean, East Asian, Middle Eastern, or Pacific Islander ancestral group, and generally is a combination of two or more of such ancestral groups.
Thus, the proportional ancestry of a test individual can include proportional affiliation among the sub-Saharan African and IndoEuropean ancestral groups (e.g., 80% sub-Saharan African and 20% IndoEuropean; or 60% sub-Saharan African, 20% IndoEuropean, and 20% of a third ancestral group); or can include proportional affiliation among the Native American and IndoEuropean ancestral groups; East Asian and Native American ancestral groups;
IndoEuropean and East Asian ancestral groups; and the lilce.
[0153] A panel of AIMs useful for estimating proportional ancestry of an individual can include AIMS as set forth in SEQ ID NOS:l to 331, for example, AIMS as set foxth in SEQ
ID NOS: l to 71, which can be useful for determining proportional ancestries including IndoEuropean, sub-Saharan African, East Asian, and Native American; or AIMS as set forth in SEQ ID NOS:7, 21, 23, 27, 45, 54, 59, 63, and 72 to 152, which can be useful for determining proportional ancestry of East Asians and~sub-Saharan Africans; or in SEQ ID
NOS:3, 8, 9, 11, 12, 33, 40, 59, 63, and 153 to 239, which can be useful for determining proportional ancestry of East Asians and IndoEuropeans; or in SEQ ID NOS:1, 8, 1 l, 21, 24, 40, 172, and 240 to 33I, which can be useful for determining proportional ancestry of IndoEuropeans and sub-Saharan Africans;.
[0154] An estimate can be made, fox example, of an individual's proportional ancestry with respect to three ancestral groups. In this method, identifying a population structure within an individual that correlates with the nucleotide occurrences of the AIMs of the test individual can be practiced by performing a likelihood determination for affiliation with each of a sub-Saharan African ancestral group, a Native American ancestral group, an IndoEuropean ancestral group, and an East Asian ancestral group; thereafter selecting three ancestral groups having a greatest lilcelihood value for the individual;
determining a likelihood of all possible proportional affiliations among the three ancestral groups having the greatest likelihood value, whereby a population structure or proportional affiliation that correlates with the nucleotide occurrences of the AIMS of the test individual is identified; and identifying a single proportional combination of maximum likelihood.
Alternatively, identifying a population structure that correlates with the nucleotide occurrences of the AIMS
can be practiced by performing six two-way (binary) comparisons comprising likelihood determinations for affiliation of each group compared to each other group;
thereafter selecting three ancestral groups having a greatest likelihood value across all comparisons;
determining a likelihood of all possible proportional affiliations among the three ancestral groups having the greatest likelihood value, whereby a population structure or proportional affiliation that correlates with the nucleotide occurrences of the AIMs of the test individual is identified; and identifying a single propoz-tional combination of maximum likelihood. Such a methodology worlcs as well for individuals of three-way admixture as individuals that are 100% affiliated with a single group.
[0155] An estimate of an individual's proportional ancestry that includes proportions of three ancestral groups also can be made by performing three three-way comparisons among the groups; determining a likelihood of all possible proportional affiliations among the three ancestral groups having the greatest likelihood value, whereby a population structure or proportional affiliation that correlates with the nucleotide occurrences of the AIMS of the test individual is identified; and identifying a single proportional combination of maximum likelihood. An advantage of the present methods is that a graphical representation of the comparison of the three ancestral groups can be generated, wherein the graphical representation comprises a triangle with each ancestral group independently represented by a vertex of the triangle, and wherein the maximum likelihood value of proportional affiliation for an individual comprises a point within the triangle (see Figures 2 and 3).
If desired, the graphical representation can further include a confidence contour that indicates a Ievel of confidence associated with estimating the proportional ancestry.
[0156] An estimate of an individual's proportional ancestry also can be made where the proportional ancestry includes proportions of four ancestral groups. In various aspects of this method, identifying a population structure that correlates with the nucleotide occurrences of the AIMS of the test individual is practiced by performing six two-way comparisons, or by performing three three-way comparisons, or by performing one four-way comparison among the groups; determining a lilcelihood of all possible proportional affiliations among the four ancestral groups having the greatest lilcelihood value, whereby a population structure or proportional affiliation that correlates with the nucleotide occurrences of the AIMS of the test individual is identified; and identifying a single proportional combination of maximum likelihood. If desired, the method can further include generating a graphical representation of the comparison of the three ancestral groups, wherein the graphical representation comprises a pyramid with each ancestral group independently represented by a vertex of the pyramid, and wherein the maximwn likelihood value of proportional affiliation for an individual comprises a point within the pyramid. If desired, the graphical representation can ftuther include a confidence contour comprising a sphere around the point, wherein the sphere indicates a level of confidence associated with estimating the proportional ancestry.
[0157] As disclosed herein, such methods are useful, for example, as a forensic tool. The present methods provide substantially greater information for forensics because, using a DNA
sample obtained at a crime scene, the methods can provide an investigator with prospective information as to the likelihood of an individuals ancestry, as well as hair, skin and eye pigmentation. In comparison, present DNA methods only allow provide retrospective information because they require that a DNA sample from a crime scene be compared with DNA samples contained in a database or taken from specific individuals: Thus, while the latter methods can provide confirmation that a suspect is likely the perpetrator of a crime, they provide no useful information until the suspect is apprehended, except in cases where the suspect's DNA sample already has been entered into a database.
[0158] The methods of estimating proportional ancestry of a test individual as disclosed herein also provide a tool that can supplement genealogical information, which generally is based on relationships established using geopolitical information (see Example 3). For example, the present methods provide information that can be used to generate an ancestral map of the world, wherein locations of populations having a proportional ancestry corresponding to the proportional ancestry of the test individual are indicated on the ancestral map. As such, the method can further include overlaying the ancestral map with a genealogical map, wherein the genealogical map indicates locations of populations having geopolitical relevance with respect to the test individual, and statistically combining the information of the ancestral map and genealogical map to obtain a most likely estimate of family history of the test individual.
[0159] Identifying a population structure that correlates with the nucleotide occurrences of the AIMs, according to a method of the invention, can be performed by comparing the nucleotide occurrences of the AIMS of the test individual with lcnown proportional ancestries coiTesponding to nucleotide occurrences of AIMS indicative of BGA. The known proportional ancestries corresponding to nucleotide occurrences of AIMs indicative of BGA
can be contained in a table or other list, and the nucleotide occurrences of the test individual can be compared to the table or list visually, or can be contained database, and the comparison can be made electronically, for example, using a computer. A
particularly useful application of a method of the invention involves associating known proportional ancestries corresponding to nucleotide occurrences of ATMs indicative of BGA of individuals, with a photograph of a person from whom the knowmproportional ancestry was determined, thus providing a means to further infer physical characteristics of a test individual. In one aspect, the photograph is a digital photograph, which comprises digital information that can be contained in a database that can further contain a plurality of such digital information of digital photographs, each of which is associated with a lcnown proportional ancestry corresponding to nucleotide occurrences of AIMS indicative of BGA of the person in the photographs.
[0160] A method of the invention can further include identifying a photograph of a person having a proportional ancestry corresponding to the proportional ancestry of the test individual. Such identifying can be done by manually looking through one or more files of photographs, wherein the photographs are organized, for example, according to the nucleotide occurrences of AIMS of the person in the photograph. Identifying the photograph also can be performed by scanning a database comprising a plurality of files, each file containing digital information corresponding to a digital photograph of a person having a lmown proportional ancestry, and identifying at least one photograph of a person having nucleotide occurrences of AIMS indicative of BGA that correspond to the nucleotide occurrences of AIMS indicative of BGA of the test individual.
[0161] According to the present invention, BGA can be determined using any of several variations of the disclosed BGA test, including three BGA tests referred to as the ANCESTRYbyDNATM 1.0 test, the ANCESTRYbyDNATM 2.0 test , and the ANCESTRYbyDNATM 3.0 test (DNAPrint genomics, Inc.; Sarasota FL), which utilize selected panel of Ancestry Informative Markers (AIMS) that have been characterized in a large number of well-defined population samples. The AIMs are selected on the basis of a showing of substantial differences in frequency between population groups and, as such, provide information as to the origin of a particular person whose ancestry is otherwise w~lcnown. For example, the Duffy Null allele (FY* 0) is very common (approaching fixation or an allele frequency of 100%) in all sub-Saharan African populations, but is not found outside of Africa. Thus, a person with this allele is very likely to have some level of African .
ancestry. Upon analysis of AIMS in a DNA sample from a person of unknown origin, a likelihood (or probability) can be determined that the person is derived from particular parental populations by calculating all of the possible mixes of parental populations. The population (or combination of populations) where the likelihood is the lughest is taken as the best estimate of the ancestral proportions of the person; confidence intervals on these point estimates of ancestral proportions are also calculated.
[0162] An objective assessment of the biological component of human ancestry provides important knowledge about the person whose DNA is examined. For example, an analysis of the biological component of ancestry can elucidate health disparities by identifying, for example, genetic contributions to the higher rates of hypertension and diabetes in African Americans, or the higher rates of dementia in European Americans. Estimates of BGA also can help connect individuals separated by adoption or some other event with their ancestral populations. And even if a person is not particularly motivated to reconnect with ancestors, lie or she can uncover the past of their family, for example, to verify family legexids or identify forgotten roots. Because the disclosed method is based on an analysis of DNA, it provides a personal demographics tool, which, unlike a census, can provide highly accurate demographics data.
[0163] There are several commercially available tests that analyze mitochondria) DNA
(mtDNA) or Y chromosome marlcers, and have been promoted as a means of learning one's ancestral origins. Although these tests can provide information regarding the provenance of some of a person's ancestors, the tests are very limited. For example, one generation ago a person has two ancestors, one mother and one father; five generations ago, a person has 32 ancestors; while 10 generations ago, a person has 1024 ancestors. Ten generations is roughly 250 years and well within the time frame of genealogical interest, especially when considering, for example, the settlement of North America. Because the mtDNA
and Y
chromosome tests only look at a small portion of the genome (the matrilineal and patrilineal lineages, respectively), they can only provide information relating to a very small proportion of a person's ancestors. The BGA test of the invention utilizes sequences throughout a person's genome and, therefore, can provide information about a greater number of ancestors.
[0164] Accordingly, the present invention provides a method of estimating, with a predetermined level of confidence, proportional ancestry of at least two ancestral groups of a test individual. Such a method, referred to as a "biogeographical ancestry test" or "BGA
test", can be performed, for example, by contacting a sample, which includes nucleic acid molecules of the test individual, with hybridizing oligonucleotides that can detect nucleotide occurrences of SNPs of a panel of at least about ten AIMS that are indicative of BGA for each ancestral group examined, wherein the contacting is under conditions suitable for detecting the nucleotide occurrences of the AIMS of the test individual by the hybridizing oligonucleotides; and identifying, with a predetermined level of confidence, a population structure that correlates with the nucleotide occurrences of the AIMS of each of the ancestral groups examined, wherein the population structure is indicative of proportional ancestry.
[0165] As used herein, the term "proportional ancestry" refers to the percent contribution of each (if more than one) ancestral group to which an individual belongs. The proportional ancestry estimated according to a method of the invention can be a proportion of any ancestral group, including, fox example, a proportion of sub-Saharan African, Native American, IndoEuropean, East Asian, Middle Eastern, or Pacific Islander ancestral group, and generally is a combination of two or more of such ancestral groups. Thus, the proportional ancestry of a test individual can include proportions of sub-Saharan African and IndoEuropean ancestral groups (e.g., 80% sub-Saharan African and 20%
IndoEuropean; or 60% sub-Saharan African, 20% IndoEuropean, and 20% of a third ancestral group); or can include proportions of Native American and IndoEuropean ancestral groups; East Asian and Native American ancestral groups; IndoEuropean and East Asian ancestral groups; and the lilce. Similarly, the proportional ancestry can include proportions of Native American, East Asian, and IndoEuropean ancestral groups; sub-Saharan African, Native American, and IndoEuropean aa.~.cestral groups; sub-Saharan African, Native American, and East Asian ancestral groups; and the like.
[0166] A panel of AIMs useful for estimating proportional ancestry of an individual can include AIMS as set forth in SEQ ID NOS:1 to 331, for example, AIMS as set forth in SEQ
ID NOS:1 to 71, which can be useful for determining proportional ancestries including IndoEuropean, sub-Saharan African, East Asian, and Native American. For example, the AIMS as set forth in SEQ ID NOS:7, 21, 23, 27, 45, 54, 59, 63, and 72 to 152 can be useful for determining proportional ancestry of East Asians and sub-Saharan Africans;
the AIMS as set forth in SEQ ID NOS:3, ~, 9, 11, 12, 33, 40, 59, 63, and 153 to 239 can be useful for determining proportional ancestry of East Asians and IndoEuropeans; and the AIMS as set forth in SEQ ID NOS: I, ~, 11, 21, 24, 40, 172, and 240 to 331 can be useful for determining proportional ancestry of IndoEuropeans and sub-Saharan Africans.
[0167] The ANCESTRYbyDNA~ 1.0 test (DNAPrint genomics, Inc.) is a first version of the BGA test that was specifically designed to provide information on the proportions of ancestry at the continental level. As such, the ANCESTRYbyDNATM 1.0 test allowed information to be obtained as to levels of Native American, European, and African ancestry, as three component groups. The ANCESTRYbyDNATM 2.0 test, in comparison, provides information on the proportions of ancestry at the continental level for most continents, including Native American, Indo-European (includes European, Middle Eastern and South Asian groups such as Indians), African, and East Asian (which includes Pacific Islanders, and can distinguish ancestries within Asia and the Pacific Rim. The ANCESTRYbyDNATM 3.0 test can further define the levels of ancestry within continents, for example, by distinguishing Japanese from Chinese, or Northern European from Middle Eastern, thus providing greater insight into where within a particular continent a person's ancestors were derived.
[0168] For the ANCESTRYbyDNATM 2.0 test, a logical grouping into four BGA
delineations was made, wherein South Asian, Middle Eastern and European are grouped into a single group called IndoEuropean (see Example 2). Tlus grouping was based on anthropological evidence and cultural connections between these groups (e.g., their languages are derived from a common base). The results disclosed herein demonstrate that these groups are far more similar to one another in genetic sequence content than to other groups. The ANCESTRYbyDNA~ 2.0 test also performs more accurately when Pacific Islanders are grouped with East Asians. As such, the four groupings used in the ANCESTRYbyDNA~
2.0 test include 1) Native American (i.e., those who migrated to inhabit South and North America); 2) IndoEuropean (Europeans, Middle Easterners and South Asians such as Indians; 3) East Asians (Japanese, Chinese, Koreans, Pacific Islanders); and 4) Africans (sub-Saharan). The ANCESTRYbyDNATM 3.0 test can further distinguish between South Asian and European, and between Pacific Islander and East Asian, thus providing 6 proportions (Native American, European, African, South Asian, East Asian and Pacific Islander), although the confidence intervals are larger than those obtained with the ANCESTRYbyDNATM 2.0 test. Further improvement to the tests axe provided, wherein the confidence intervals are reduced. Confidence intervals around a point estimate can be reduced, thus increasing the accuracy of the test, by analyzing a complementary panel, thereby improving the confidence intervals by about 50%.
[0169] The algorithm used to determine the ancestral proportions was developed based on the idea that it is possible to use certain statistical methods to make an inference of the proportionality of ancestry in an individual sample based on their sequence (see Example 6;
see, also, Table 12). The method of making this inference using the present algorithm is similar to those of others, wherein, if the frequency of an allele in a population is known, and this frequency is significantly different from population to population, a "Maximum Lilcelihood Estimation" (MLE) can be used to determine the probability that a person with the allele belongs to one of the groups. Expanded to include multiple alleles from multiple genetic loci and multiple populations, the process is the same. By way of simplification, Bayes' theorem states that the probability of an event given a circumstance (called a posterior probability) is a function of the frequency of the circumstance given the event (a conditional probability) and the frequency of the event itself (the prior probability). By determining the probability of the event given the circumstance for a wide range of possible events, that with the highest probability can be selected, thus obtaining the MLE for the probability.

[0170] In the present algoritlnn, the event is a proportionality of ancestry, and the circumstance is the genotype of the individual. If the minor allele frequency for 10 SNPs in 2 populations of human beings is known, and the sequence of a person at each of the 10 SNPs is known, a simple binary classification into one of the two groups can be made by choosing that for which the conditional probability is highest. This would offer little improvement over current methods for determining the BGA from a DNA sample. What is provided by the present invention is the ability to obtain the proportionality of ancestry for more complex and realistic scenarios of ancestry. There axe many possible combinations such as 99%, African, 1 % European, 0 Native American, 0% East Asian; or 98% African, 1 % European, 1 % Native American, 0% East Asian; and the like. The posterior probabilities for each of the thousands of possibilities are not the same for any particular individual, given his or her multilocus genotype (i.e., genotypes of many AIMs), and, in fact, there is one that has the highest posterior probability or likelihood for each genotype. It is this combination that the present algorithm selects (i.e., the MLE).
[0171] Previous methods have been limited in that the confidence of the estimate was not known. The present algoritlnn addresses this limitation by plotting the MLE
graphically, including plotting the confidence regions around the MLE such that a level of confidence can be ascertained (see Figures 2 and 3). Further, the algorithm (i.e., the software code) that performs the MLE calculation operates in an unusually efficient manner. The triangle plot provided by the algorithm is an original method to graphically represent the MLE
calculations and their confidence intervals. To read a triangle plot (see below), a perpendicular line is dropped from each vertex (triangle point) of the triangle to the opposite edge (base) of the triangle (see Figure 2A). In this figure, the circle represents the MLE, and a line has been dropped from the Native American (NAM) vertex to the line below; the line serves as a scale for the percentage of Native American ancestry, from 0% at the base to 100% at the vertex (or tip). Projecting the circle on this line can be analogized to holding a flashlight to the right of the triangle at the same level as the circle and observing the shadow the circle makes on the line. Where this shadow falls on the line indicates the percentage of Native American ancestry. In this example, the individual is about 15% Native American, as indicated by the hash mark on the line.

[0172j The results provided using the disclosed method provide a statistical estimate of BGA admixture for an individual (the Maximum Likelihood Estimate (MLE)), which is indicated as a point on a triangle plot to represent the proportions of the most relevant three groups for the individual. While the MLE is the most likely estimate, the true value for the individual can be a different set of proportions. A triangle plot with calculated and plotted estimates that are 2 times, 5 times and 10 times less likely than the MLE is exemplified. The first contour around the MLE delimits the space within which the estimates are up to 2 times less likely, with those positions near the line reflecting values close to 2 and those near the MLE closer to 1; the second contour around the MLE delimits the space within which the estimates are 5 times less likely in the same graded fashion proceeding from the first contour line to the second contour line. The third contour delimits the space within which the estimates are from 5 fold (near the second contour line) up to 10 times less likely (near the third contour line). The greater the number of DNA positions read, the closer these contour lines approach the MLE point. On the triangle plot, the likelihood (probability) that the true value is represented by a different point than the MLE increases until the MLE
is met, where the probability is maximum (i.e., the Maximum Likelihood Estimate; MLE). The test can be performed so that the contour lines are very close to the MLE by sequencing a very large collection of markers. However, to beep the test affordable and efficient, the survey can be limited to a desired number of markers (e.g., 10, 15, 20, 25, 30, 40, 50, 60, 70,80, 90, 100, or more) that is sufficient to determine the most likely proportions with good confidence. In this respect, a variety of different panels of 100 SNP markers have been examined, a panel of 71 AIMS has been used in a number of studies, and a panel of 175 AIMs is being examined such that very confidence is achieved.
[0173] The BGA test of the invention has been validated by determining the frequency of DNA sequence variants in various human populations. In addition, the test has been evaluated using a large number of people from a wide range of ancestral groups, and the estimates have corresponded well to what is lcnown from anthropological and historical data.
For example, Hispanics are known to have arisen as an ethnic group from the blending of colonial Europeans with Native Americans, and the hundreds of Hispanics examined using the BGA test aligned with these two groups almost exclusively. As another example, though Nigerians plot as of almost ptue African BGA, African Americans plot more as a mixture between this group and Europeans, which is what would be expected from knowledge about the admixture between Africans and Europeans in the United States.
[0174.] The method also was validated through pedigree challenge (see Example 1); i.e., when the BGA is determined from a mother and father, that of their children should plot somewhere between the two. Numerous family pedigrees have been examined using the test, and the ancestral proportions of offspring have always plotted between those of the child's parents. When the MLE estimates are tested objectively (blindly), they prove to be excellent estimates of axlcestral proportions. For example, the data for a European American man, whose mother is European mix and father is mostly Greek, showed the man to be of 85%
European ancestry, but also of 15% Native American ancestry (Example 1). In fact, his paternal great-grandmother was full-blood Cherokee, thus confirming the result of the test (based on the laws of genetics, the man would be predicted to have a 12%
Native American Ancestry if his great-grandmother was 100% Native American and none of his other relatives were of Native American ancestry). Further, the man's wife is Mexican, and she was determined to be of mostly Native American, but with some Native American and African heritage. This was also expected based on what is known from anthropological origin of Hispanics, who derived from the union of Spanish explorers with Native Americans in Colonial Caribbean and Latin America. Each of the three children of the man and woman plotted roughly half way between both parents, as expected. None of the children showed any Asian or Pacific Islander ancestry, which would have been impossible (assuming an accurate test) because none of the parents showed any significant Asian or Pacific Islander heritage, and none of the children were found to have more African ancestry than their mother, which would also be impossible given the fact that the father has virtually none.
Thus, the results of the children were consistent with those of the parents, and the MLE
values were accurate estimates when tested against what was known from biographical data.
[0175] The genotypes (nucleotide letters) determined to date are quite accurate. Because the latest genetic reading equipment available is used, an accuracy greater than 99% accuracy is routinely achieved for each site. If an accurate value was not obtained for a particular site in a particular sample, an "FL" is indicated, instead of the genotype letters for that site.
Having a few FL's generally does not prevent a good ancestry estimate. A
sample can produce an FL for a site because, for example, a small region of the chromosome around this site is missing or is of different sequence character than for most (this result is not uncommon given the highly variable nature of the chromosomal positions we measure); or because not enough DNA was obtained from the buccal swab used to collect a DNA sample.
[0176] The genome was scanned for a useful panel of BGA AIMS and the best 71 AIMS
were identified using the maximum likelihood algorithm to measure BGA
admixture proportions were selected (Table 6). Using these AIMS, majority BGA
affiliations were measurable in a manner consistent with self held notions on BGA, and BGA
admixture proportions were measurable with significantly improved precision, accuracy and reliability compared to previously described methods for the inference of race (see Example 2; see, also, Example l, using 32 marker test). This test can be used during study design to help reduce or eliminate the insidious effects that cryptic or micro population structure imposes. The test also can be useful for forensic scientists who currently use imprecise and sometimes inaccurate means by which to infer race from crime scene DNA.
[0177] The present invention also provides articles of manufacture, including one or a plurality of photographs, each photograph being of a person having a lazown proportional ancestry corresponding to a population structure comprising nucleotide occurrences of AIMS
indicative of BGA, the known proportional ancestry being associated with the photograph in the article. An article of manufacture of the invention (i.e., a photograph and the proportional ancestry information) can be contained in one or more files (e.g., the photograph and information in one file, or the photograph in one file and the information in a second file, which is or can be linked to the photograph). If desired, more than one photograph of an individual having a known proportional ancestry can be contained in the same or a linlced file, for example, photographs containing different profiles of the individual or photographs of the individual at various ages.
[0178] Similarly, a plurality of the articles (i.e., photographs and proportional ancestry information) can be contained in a file, for example, a file containing a plurality of photographs of different persons, wherein the some or all of the persons have the same or different known proportional ancestries that correspond to a population structure comprising nucleotide occurrences of AIMS indicative of BGA. Such a plurality of articles also can be contained in different files, including, for example, a plurality of files, each containing one photograph and information regarding the knovcm proportional ancestry of the individual in the photograph, or each containing two or more photographs of different individuals, each of which contains the same lcnowwnn proportional ancestry, or each containing two or more photographs of different individuals, some or all of which have a different proportional ancestry as compared to another individual whose photograph is contained in the file.
Accordingly, a plurality of such articles is provided, as is a plurality of files, each file of which can contain one or more articles, i.e., photographs, which can be of one or more persons having the same or different lcnown proportional ancestries that correspond to a population structure comprising nucleotide occurrences of AIMS indicative of BGA; and the plurality of files can contain files, each of which contains one or more photographs of one or more persons, and when containing one or more photographs of two or more different persons, the different persons can have the same or different known propoz-tional ancestries.
[0179] The article of manufacture, i.e., the photograph of a person having a known proportional ancestry corresponding to a population structure comprising nucleotide occurrences of AIMS indicative of BGA can be a digital photograph, which comprises digital information, including for the photographic image and any other information that may be relevant or desired (e.g., the age, name, or contact information of the subject in the photograph, or the subject's answer on a questionnaire as to what the subject believes his or her ancestry to be). Such digital information of one or more digital photographs can be contained in a database thus facilitating searching of the photographs and/or known proportional ancestry information using electronic means. As such, the present invention further provides a plurality of the articles of manufactures, including at least two digital photographs, each of which comprises digital information. Where the digital information for one or a pltuality of the articles is contained in a database, it can comprise a,ny medium suitable for containing such a database, including, for example, computer hardware or software, a magnetic tape, or a computer disc such as floppy disc, CD, or DVD.
As such, the database can be accessed through a computer, which can contain the database therein, can accept a medium containing the database, or can access the database through a wired or wireless network, e.g., an intranet or Internet.
[0180] The present invention also provides lcits useful for practicing a method of the invention. Such kits can contain, for example, a plurality of hybridizing oligonucleotides, each of which has a length of at least fifteen contiguous nucleotides of a polynucleotide as set forth in SEQ ID NOS:I to 331 (or a polynucleotide complementary thereto), the plurality including at least five (e.g., 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, etc.) of such oligonucleotides, each based on different polynucleotides as set forth in SEQ ID NOS:1 to 331.
In one embodiment, the hybridizing oligonucleotides that include at least fifteen contiguous nucleotides of at least five polynucleotides as set forth in SEQ ID NOS:1 to 71; or polynucleotides complementary to any of SEQ ID NOS:1 to 7I. In another embodiment, the hybridizing oligonucleotides are specific for at least ten AIMS as set forth in SEQ ID NOS:1 to 71. A kit of the invention also can contain at least two panels of such hybridizing oligonucleotide, including, for example, a panel of at least five (e.g., 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, etc.) hybridizing oligonucleotides specific for AIMS as set forth in SEQ ID
NOS:7, 21, 23, 27, 45, 54, 59, 63, and 72 to 152; or a panel of at Ieast five hybridizing oligonucleotides specific for an AIM as set forth in SEQ ID NOS:3, 8, 9, 1 l, 12, 33, 40, 59, 63, and 153 to 239; or a panel of at least five hybridizing oligonucleotides specific for AIMS
as set forth in SEQ ID NOS:1, 8, I 1, 21, 24, 40, 172, and 240 to 331; or two or more of such panels and/or a panel of at least five hybridizing oligonucleotides specific for AIMs as set forth in SEQ ID NOS:1 to 71.
[0181] The hybridizing polynucleotides of a Icit of the invention can include probes, which are useful for detecting a particular AIM, including a particular nucleotide occurrence at the SNP position of the AIM; can include primers, including primers useful for a primer extension reaction and primer pairs useful for a nucleic acid amplification reaction; or can include combinations of such probes and primers. A hybridizing oligonucleotide of the plurality can, but need not, include a nucleotide corresponding to nucleotide position of the SNP or DIP of an ATM, e.g., nucleotide 50 of an AIM as set forth in any of SEQ
ID NOS:l to 55 and 57 to 331 or nucleotide 26 of SEQ ID N0:56, or to a nucleotide sequence complementary thereto, such a hybridizing oligonucleotide being useful as a probe to identify the presence or absence of a particular nucleotide occurrence at the SNP
position of the AIM.
[0182] A lcit of the invention also can contain at least one pair of hybridizing oligonucleotides useful for detecting the nucleotide occurrence at the SNP
position or the presence or absence of a nucleotide sequence the DIP position of an AIM. For example, a pair of hybridizing oligonucleotides can include one oligonucleotide that hybridizes upstream and adjacent to the SNP position of an AIM and a second oligonucleotide that hybridizes downstream of and adjacent to the SNP position of the AIM, wherein one or the other~of the pair further contains a nucleotide complementary to a nucleotide occurrence suspected of being at the SNP position of the AIM (i.e., one of the polymorphic nucleotides), such a pair of hybridizing oligonucleotides being useful in an oligonucleotide ligation assay. In another example, a pair of hybridizing oligonucleotides can include an amplification primer pair, including a forward primer and a reverse primer, such a pair of hybridizing oligonucleotides being useful for amplifying a portion of polynucleotide that includes the SNP
or DIP position of the ATM.
[0183] A kit of the invention can further contain additional reagents useful for practicing a method of the invention. As such, the lcit can contain one or more polynucleotides comprising am AIM, including, for example, a polynucleotide containing an AIM
for which a hybridizing oligonucleotide or pair of hybridizing oligonucleotides of the lcit is designed to detect, such polynucleotide(s) being useful as controls. Further, hybridizing oligonucleotides of the lcit can be delectably labeled, or the lit can contain reagents useful for detectably labeling one or more of the hybridizing oligonucleotides of the lit, including different detectable labels that can be used to differentially label the hybridizing oligonucleotides; such a lcit can further include reagents for linking the label to hybridizing oligonucleotides, or for detecting the labeled oligonucleotide, or the like. A kit of the invention also can contain, for example, a polymerase, particularly where hybridizing oligonucleotides of the lcit include primers or amplification primer pairs; or a ligase, where the kit contains hybridizing oligonucleotides useful for an oligonucleotide ligation assay. In addition, the lcit can contain appropriate buffers, deoxyribonucleotide triphosphates, etc., depending, for example, on the particular hybridizing oligonucleotides contained in the lcit and the purpose for which the lcit is being provided.
[0184] The following examples are intended to illustrate but not limit the invention.

DETERMINATION OF BIOGEOGRAPHICAL ANCESTRY
USING ANCESTRY INFORMATIVE MARKERS
[0185] This Example demonstrates that a panel of 32 Ancestry Informative Markers (AIMS) allows an estimate of the genetic contribution from populations of African, European and Native American ancestry.
[0186] The AIMS used in the exemplified study include single nucleotide polymorphisms (SNPs), deletion/insertion polymorphisms (DIPs) and Alu sequences (see Example 2 for identification of AIMs). Markers showing differences between the parental populations greater than 30% were selected (Table l; see, also, SEQ ID NOS:332-363).
Informative genetic markers were identified by testing each candidate marker in a panel of European (Spanish, and German), African (from Nigeria, Sierra Leone, and Central African Republic), and Native American populations (Mayan and South Western Native Americans) to confirm the usefulness of the marlcer for admixture estimation.

Ancestry Informative Marker Panel MARKER LOCATION Mb AF/EU AF/NA EU/NA

MID 575 (356*)1p34.3 ~42 O.I30 0.417 0.546 MID 187 (357)1p32 50.2 0.370 0.440 0.070 FY-NULL (339)1q23.2 181 0.999 0.999 0.000 AT3 (359) 1q25.1 196 0.575 0.777 0.202 F13B (338) 1q31.3 220 0.641 0.674 0.033 TSC1102055 1q32.1 234.5 0.441 0.303 0.744 (343) WI-11392 (NS**)1q42.2 269.5 0.444 0.256 0.188 WI-16857 (345)2p16.1 56.2 0.536 0.548 0.012 WI-11153 (346)3p 12.1 95.0 0.652 0.022 0.629 GC*IF (NS) 4q13.3 75.7 0.697 0.530 0.166 GC*1S (NS) 4q13.3 75.7 0.538 0.478 0.060 MID-52 (NS) 4q24 110.7 0.186 0.500 0.687 SGC306I0 (354)Sq11.2 61.5 0.146 0.281 0.427 SGC30055 (355)Sq22.1 124.7 0.457 0.675 0.218 WI-17163 (347)Sq33.1 173.9 0.120 0.641 0.521 WI-9231 (348)7p22.3 ~1.2 0.017 0.387 0.370 WI-4019 (349)7q21.3 100 0.124 0.173 0.296 CYP3A4 (NS) 7q22.1 101.9 0.761 0.755 0.006 LPL (340) 8p21.3 22.3 0.479 0.521 0.042 CRH (NS) 8q13.2 73.2 0.609 0.655 0.046 WI-11909 (350)9q21.3I 81.0 0.075 0.587 0.663 DIIS429 (337)l lq13.3 70.4 0.429 0.054 0.376 TYR (344) l lq14.3 95.4 0.444 0.055 0.389 DRD2-Taq I l 1q23.2 125.0 0.535 0.046 0.582 "D"

(336) DRD2-Bcl I 11q23.2 125.0 0.080 0.565 0.485 (335) APOA1 (360) 11q23.3 128.9 0.505 0.555 0.050 GNS3 (332) I2pI3.31 ~7.2 0,463 0.430 0.033 RB1 (361) 13q14.2 47.4 0.611 0.711 0.100 OCAZ (342) 15q12 24.0 0.631 0.369 0.263 WI-14319 15q14 30.0 0.185 0.310 0.494 (351) CYP19 (334) 15q21.2 47.6 0.045 0.379 0.423 PV92 (362) 16q23.3 96.5 0.073 0.551 0.624 MC1R314 (341)16q24.3 103.8 0.350 0.441 0.090 WI-14867 17p 13.2 ~3 .5 0.448 0.404 0.045 (352) WI-7423 (353)17p 13.1 ~8.2 0.476 0.074 0.402 Sb19.3 (363)19p13.11 27.0 0.488 0.236 0.253 CKM (333) 19q13.2 55.8 0.150 0.694 0.545 MID 154 (358)20q11.23 ~34 0.444 0.368 0.076 MID 93 (NS) 22q13.2 38.6 0.554 0.179 0.733 ~

Shown are the marker name axed chromosomal band, approximate location of the marker on the chromosome in megabases (Mb), and difference in frequency between African and European populations (AF/EU), African and Native Americans (AF/NA) and European and Native Americans (EU/NA). Differences greater than 30% are marked in bold letters (see, also, Shriver et al., s~rp~a, 2003, which is incorporated herein by reference) * Numbers in parentheses are SEQ ID NO: for AIMS; NS - sequences not shown.
[0187] The publicly available human genome sequence database and polymorphism database were screened in order to identify SNPs that met the criteria for being a good AIM.
Allele frequencies are available for many of the SNPs in the public databases for three populations - Africans, Europeans and Asians. Since these frequencies are obtained from small samples they are not always accurate. The main criteria for selection herein was the delta value that derived from using these frequencies, which is a statistical measure of the difference in minor allele frequency between various populations of human beings. For example, a C or a G polymorphism at a particular place in the human genome, where the C is present mainly in individuals of European descent and the G present mainly in individuals of Native American descent, would have a high delta value and, therefore, qualify as a good AIM. Similarly, an A or a C polymorphism at a particular place in the human genome, where the A is present mainly in individuals of African descent and the C present mainly in individuals of Asian descent would have a large frequency differential between these groups and, therefore, a high delta value, thus qualifying as a good AIM. A list of such "candidate AIMS was compiled, ranlted from laxgest delta value to smallest delta value for each of the possible pair-wise population comparisons, and screened, one at a time, against a panel of "parental" samples. Parental samples are samples from regions of the world that are relatively homogeneous, for example, Niger or Congo for sub-Saharan Africans, Southern Mexico for Native Americans, China for East Asians, and Europe for Europeans.
[0188] About half of the candidate AIMS proved to be not very useful because their actual delta values were not as high as expected from the public database allele frequencies (some were not even SNPs, or could not be assayed using the present platform).
Sequences that were validated as true AIMs, such as those exemplified herein, were useful for admixture mapping, malting inferences of individual ancestry proportions, and malting inferences of population group admixture proportions, as well as for screening genomes in order to identify marlters with alleles that correlated with certain human traits through their ancestry informativeness. Even though each candidate ATM was initially selected from the public database based on crude population structure differences (i.e., continental populations), many of them were found to carry information on finer levels of structure because the separation of subgroups of humans from larger groups throughout human evolution has provided a fertile opportunity for genetic drift, founder effects, and natural selection to operate in either fixing or eliminating their sequences.
[0189] Sequences are shown in the Sequence Listing from 5' to 3' (left to right), and, for SEQ ID NOS:1 to 33I, with the SNP generally, but not always, at nucleotide position 50 from the 5' terminus (except for SEQ ID N0:56, position 26). The polymorphism is indicated with an IUB symbol, wherein S=C/G, Y=C/T, R=A/G, I~=G/T, W=A/T, etc.
As such, the disclosed sequences (SEQ ID NOS:1 to 331) provide information as to the target being examined (i.e., the polymorphism) as well as information for preparing primers and amplification primer pairs, and hybridization probes, for sampling the SNP
(i.e., determining the genotype of a sample). Further, the disclosed sequences can be used, if desired, to scan public databases to identify additional upstream and downstream nucleotide sequences.
[0190] This panel of markers was extremely powerful for estimating with precision admixture proportions in population samples (standard error typically between 1% and 5%).
In addition, the AIMS provided reasonable estimates of individual ancestry, and suggested that an equivalent precision can be obtained using more markers (confirmed in Example 2).
Two independent methods, a Maximum Lilcelihood Estimate (MLE) method (Chakraborty et al., supra, 1986) and a Bayesian method using the program STRUCTURE (Pritchard et al., supra, 2000) estimating individual ancestry, were used; the values obtained by both methods were highly correlated (R2 = 0.9836) when estimates of individual ancestry were compared in terms of percent African genetic contribution in a sample of African Americans from Washington DC. These marlcers are excellent for determining whether there is population structure in a sample from an admixed population. This ability is important in terms of admixture mapping applications because, as discussed below, the process of admixture can produce significant structure in a population, and consequently a high number of false positive results (positive associations caused by genetic structure, not by physical linlcage of a marker with a disease causative gene), increasing significantly the risk of misinterpreting mapping results.
[0191] The present study confirms that AIMs can be identified using the disclosed methods, and provides a panel of 32 AIMs that can be applied towards an ultimate goal of compiling a panel of approximately 1,000 AIMs spanning the entire human genome.
Candidate AIMs were obtained by screening SNP allele frequency data generated through The SNP Consortium (TSC). Six sites, including the Sanger Centre, Celera Genomics, Washington University, Orchid Biosciences, Motorola, and Whitehead lizstitute, have generated, as of 2003, allele frequencies on 60,000 SNPs located throughout the genome using a central collection of 42 individuals from each of 3 populations (African-American, European-American, and Asian-American). This database, which is freely available to researchers (see, e.g., using hypertext transfer protocol ("http"), at URL
"snp.cshl.org"), has been used to provide the present results, thus demonstrating the usefulness of the resource to compile a genome-wide panel of AIMS.
[0192] The present study focused on the accuracy of the SNP database and the number of candidate SNPs present therein. With respect to the accuracy of the database, each group involved in the SNP consortium has taken a different approach to generating data. As such, initial concerns regarding how the data can be combined was addressed. Because the genotyping approaches were different for each group, it was necessary to address the question of ascertainment biases that might differentially affect the data of particular groups. For example, most of the groups produced their allele frequencies after sequencing a subset of the TSC diversity panels, then scoring these markers in the larger groups of 42 individuals from the 3 populations. The Washington University group has taken an approach whereby pooled sequencing throughout regions was performed, and the allele frequencies calculated for variable positions discovered during this effort. The Orchid group has not used sequencing but, instead, stat-ted with loci from the TSC SNP database that are known to be polyrnorphic.
Given such differences, a systematic characterization was made as the extent, if any, that different biases may have affected the results.
[0193] One approach for systematically characterizing such potential bias was to compare the allele frequencies for loci that were genotyped by more than one group.
Although there was dispersion around the 45° line, as expected, there was general agreement in the frequency data obtained by the different groups (R2=0.8762), indicating that the extent of the allele frequency bias introduced by the differing genotyping and ascertainment strategies was limited. The next step in testing the accuracy of these data was to separate the data by site and perform pair-wise comparisons, which would allow the identification of particular sites that have allele frequency estimates that deviate more when compared to other sites.
[0194] With respect to the number of candidate AIMs, it also was important to determine how many of the 60,000 SNPs characterized by the TSC would be useful for admixture mapping. Since it can be useful ultimately to compile a panel of about 1,000 markers showing large frequency differences between the relevant population groups (Fst>0.4), it was important to evaluate which percentage of the available markers have the desired characteristics. Candidate AIMS were based on the recommendation of McI~eigue et al.
(supra, 2000). The markers with information available for African, Asian and European populations, the cumulative proportion of markers in each Fst category (0-1, in 0.05 intervals) and the total number of candidate ATMs for each possible comparison. The distribution of pairwise Fst from the TSC allele frequency project was as follows: Asian-European (556 candidate AIMs/25,110 total SNPs; average Fst = 0.0720); Asian-African (1026 candidate AIMs/25,578 total SNPs; average Fst = 0.0886); and European-African (1306 candidate AIMs/30,103 total SNPs; average Fst = 0.0861). As such the screen revealed that about 2-5% of the markers can be useful for admixture mapping [0195] The geographical pattern of admixture in the US admixed populations, particularly African Americans and Hispanics, was the subject of an initial examination.
The admixture proportions of more than I 8 African-American populations were characterized, and a map was generated showing an estimate of the European genetic contribution to African Americans from several different geographic areas in the United States. The European admixture ranged between 3.5% in the Gullah of South Carolina to 22.5% in New Orleans (e.g., 18.8% in Chicago; and 16.4% in Houston). Most of these estimates were obtained using an initial panel of 10 infomnative AIMS. The observed distribution was interpreted in terms of well knovnm historical and demographic events that have played an important role in African American history (see Parra et ah., supra, 1998, Parra et al., supy~a, 2001 ). These data allow the application of admixture mapping to identify genes involved in complex diseases.
It is expected that admixture mapping will be more suitable in populations showing a high degree of admixture and, therefore, populations such as the Gullah (3.5%) and Jamaicans (6.6%), in which European genetic contribution has been very limited, may not be suitable for this kind of analysis.
[0196] In preliminary studies using mitochondrial DNA (mtDNA), it was observed that African Americans have a detectable, although how, Native American genetic contribution, in accordance with the self reported Native American ancestry often mentioned by African-American individuals. Having identified 30 AIMS informative for African/Native American and I9 AIMs for European/Native American contrasts (see Table 1; see, also, Shriver et ah., sups~a, 2003), the presence of Native American admixture in three African-American populations was examined using nuclear DNA markers. In accordance with the mtDNA
estimates (which only provide "maternal" contribution information), evidence of a low Native American genetic contribution was detected in each of the African-American samples (Washington DC, 6%; Afro-Caribbeans from London, 5%; and Bogalusa, Louisiana, 6%).
[0197] Regarding admixture in Hispanics, the relative European, Native American and African contribution in a sample of Spanish-Americans from San Luis Valley CO
was estimated. A 59% European admixture, 35% Native American admixture, and 6%
African admixture was observed in this sample, in good agreement with estimates previously described for populations of Mexican ancestry (Chahcraborty et al., sup~~a, 1986; Hanis et al., sups~a, 1991; Tseng et al., A~ae~. J. Phys. A~cthf opol.. 106:361-71.1998;
Collins-Schram et al., supra, 2002). As shown in Example 2, further characterization of admixture in additional samples from Mexico and in two samples from Hispanics of Puerto Rican ancestry (New York and Puerto Rico) has been performed.
[0198] A sample of individuals of European ancestry (N=199) currently living in State College PA also were analyzed. The genetic contribution in this sample was predominantly of European origin (91%), with evidence of some African (3%) and Native American (6%) influence. These results are summarized in Figure 4, using a triangular plot, which clearly reveals the differences in average admixture levels between European Americans, Spanish Americans, and African Americans. The triangular plot shown in Figure 4 represents the average admixture estimate in a particular sample; the underlying distributions of individual ancestry are complex, with different individuals showing widely dispersed values of African, European and Native American ancestry (not shown). Tn African Americans, most individuals showed predominantly African genetic contribution, but some persons showed relatively high European contribution and also, to a lesser extent, Native American ancestry.
European Americans clustered more tightly near the pole corresponding to high European contribution, with few persons showing evidence of Native American and African ancestry.
Spanish Americans showed the highest dispersion of individual ancestry, as expected given the high admixture level observed in this sample.

[0199] Notably, individuals showed the whole range of European and Native American ancestry (from 100% European to 100% Native American), and a relatively lower African genetic contribution also was evident in some individuals. Some of the variance observed in individual ancestry was likely due to the stochastic error due to the limited number of markers used to infer ancestry. Thus, while the 20-32 markers used in the exemplified test detected individual ancestry, the standard error of the estimates was fairly high; increasing the number of AIMs is expected to increase the precision of the individual ancestry estimates (see Example 2). The other component of the variation in individual ancestry was due to true differences in ancestry between individuals. The remarlcable correlation in individual ancestry values obtained by two totally independent methods, ML and STRUCTURE
(discussed above), indicates that this panel of markers can capture the underlying individual ancestry patterns characteristic of these populations. As disclosed below, controlling for variations in individual ancestry allows the avoidance of false positive results.
[0200] The effect of admixture dynamics in population structure and the extent of linkage disequilibrium (LD) was examined. The importance of the admixture model (hybrid isolation model vs. continuous gene flow model) in terms of the population structure and LD created in the admixture process was previously described (Pfaff et al., supra, 200I), and two methods to quantify the level of population structure in admixed populations were presented.
Population structure is a lcey aspect of admixture mapping, as well as of any genetic association study in an admixed population. This issue has been explored in an African-American, a Spanish-American and a European-American sample using more informative marlcers than previously available.
[020I] The presence of structure was evaluated in two different ways. First, the observed number of significant associations was compared between unlinked markers with the number expected at the 5% significance level, and second, the average correlation of individual ancestry was estimated using two subsets of genetic markers. In agreement with previously reported data, the African-American population from Washington DC showed a significant genetic structure, as reflected by a much higher number of significant associations between unlinked markers than expected by chance (10.5% vs. 5%, Figure SA). Very strong associations were observed between markers located as far apart as 24 Mb (AT3-F13B, G=15.21, p<0.0001), providing clear evidence that these significant associations are caused by the admixture process. The alleles that were associated were always combinations of the alleles that are frequent in African populations, with the higher the difference in frequency, the more frequently associations were observed between markers: FY, showing the highest difference in frequency between African and Europeans, was significantly associated with 9 unlinked marlcers. Thus, the admixture process in tlus African-American population, showing 17% European ancestry, created a strong association between markers showing high frequency differences between African and European populations. Although these associations are significant both when the markers are linked and when they are unlinked, linked markers tended to show higher G values than unlinked markers, indicating that the association due to true linkage can be distinguished from the association due to genetic structure, as previously demonstrated (McKeigue et al., supra, 2000).
Interestingly, significant associations were observed between marlcers showing high frequency differences between European and African populations, but not between markers showing high frequency differences between African and Native American, or European and Native American populations. This result was likely due to the low Native American ancestry observed in this sample (6%), which was insufficient to create detectable associations due to the admixture process in such a small sample.
[0202] Another line of evidence demonstrating the high level of genetic structure present in this African-American sample was the signficant correlation between independent estimates of individual ancestry using different subsets of the genetic markers. The average correlation over 100 random selections of independent subsets of markers was r=0.40, p<0.0001 (Figure SB). The pattern of genetic structure and LD observed in Waslungton DC
African Americans and in other African-American samples analyzed with a more limited set of markers (Jaclcson MI, and the Low Country counties, South Carolina, Pfaff et al., supra, 2001) indicated that the best model to describe the admixture process in these populations was the continuous gene flow model. Additional support for this model came from the strong correlation between Do (the initial expected association between markers originated by the admixture process) and Dt (the current association between markers). As shown using computer simulations (Pfaff et al., supf°a, 2001), in populations following the continuous gene flow model, positive correlations between Do and Dt are expected and, in fact, this result was observed in African-American populations. In populations following the hybrid isolation model, significant coiTelations between Do and Dt are not expected.
[0203] The Spanish-American sample from San Luis Valley that was analyzed showed less genetic structure than any of the African-American populations. The number of observed significant associations between unlinked rnarlcers was only slightly higher than expected at a 5% significance level (7.3% vs. 5%, Figure SA). This result was interesting considering that the Spanish-American population was considerably more admixed than the African Americans, and under the same model of admixture dynamics, would be expected to show considerably more structure. The correlation of individual ancestry estimates based on independent markers, although significant, was much lower than the values observed in the African-American populations (r=0.11, p<p,0001, Figure SB). Also, there was no correlation of Do and Dt in the San Luis Valley sample, in contrast to the results observed in African Americans. These results demonstrate that admixture dynamics (the way in which the population was formed and has evolved) have been different in the African-American populations and the San Luis Valley population, with the former more closely resembling the continuous gene flow model than the latter. Of course, other Hispanic populations can show different patterns of admixture dynamics than that observed in San Luis Valley.
[0204] As expected from the lower admixture levels observed in European Americans, there was no evidence of genetic structure due to admixture in this sample from State College PA (Figures SA and SB). The number of significant associations between unlinced markers was similar to the value expected by chance, and there was no correlation between individual ancestry estimates of independent subsets of markers (p=0.149, NS).
[0205] These results demonstrate that the use of selected genetic markers (AIMS) allows an analysis of the dynamics of the admixture process and the effect of this admixture process on the pattern of LD in admixed populations. In admixed populations that have had an admixture process similar to the hybrid isolation model (initial admixture followed by independent evolution of the admixed population without further genetic contribution of the parental populations), few false positive results are expected (recalling, that the falseness is relevant for a "gene hunter" searching for genes that cause traits through LD
or linkage, not for those seelcing to develop classification tools). In admixed populations that more closely resemble the continuous gene flow model (continuous genetic contribution each generation from one of the parental populations to the admixed population), the LD is expected to extend much longer distances and problems with false positive results will arise.
Fortunately for the gene hunter, the information conveyed by the AIMS can control for genetic structure and minimize false positives. An example is provided below demonstrating how such control can be achieved using appropriate statistical methods and skin pigmentation as a model phenotype. ' [0206] Skin pigmentation and individual ancestry was examined in an African-American sample and a Spanish-American sample. As previously demonstrated, the genetic structure created by admixture can be effectively controlled, and association due to linkage can be distinguished from spurious association due to genetic structure using appropriate statistical tests (McKeigue et al., supra, 2000). In the present study, the same methods were applied in a study of skin pigmentation in two admixed samples (African Americans from Washington DC and Spanish Americans from San Luis Valley). Information on slcin pigmentation was collected for each individual in both studies, and the subjects were genotyped for a panel of ATMs and individual ancestry proportions were calculated using the maximum lilcelihood method (Chalcraborty et al., supra, 1986). Individual ancestry (% African or %
Native American) was plotted against melanin index (African) or skin reflectance (Native American) for each individual. Several of the AIMS showing high differences in frequency between the parental populations were also candidate genes for pigmentation.
[0207] In the African-American sample, a strong and highly significant correlation (R2=0.1879, p<0.0001) was observed between individual ancestry and the melanin index, which measures the melanin content of the skin. Individuals with darker skin had, on average, higher levels of African ancestry. The individual ancestry estimates were based on 21 markers, and, therefore, subject to a relatively high variance, thus explaining at least some of the dispersion observed in the graph. An interesting feature of these results was the evident decrease in variance observed in moving from the right (more African ancestry) to the left (more European ancestry). This result is consistent with the higher level of variability in skin color that is found in African populations as compared to Europeans.
The high correlation observed between individual ancestry and slcin pigmentation can be due to the population structure typical of African-American populations (as discussed above), and related to the limited munber of genes that were used for determining the parental population differences contained within this relationship.
[0208] A similar plot was prepared for the San Luis Valley sample. Individual ancestry estimates using 15 Native American/European AIMS were plotted against pigmentation level as measured by the percent of light reflected through the PHOTOVOLT 670 Green filter.
Because slcin pigmentation was measured in different ways (absorbance vs.
reflectance) in these two studies, the trends observed when graphed are reversed. In the Spanish-American sample, the correlation between individual ancestry and skin color also was significant (R2=0.0481, p<0.001), but lower than in the African-American sample, possibly due to the reduced genetic structure present in this sample.
[0209] Tests for differences in the average pigment levels by genotype for the AIMs typed in the African-American population sample discussed above were performed. The panel of AIMS included three candidate gene markers, OCA2, TYR, and MC1R. The analysis was performed in three alternative ways: first with no consideration of the individual ancestry estimates (ANOVA); second after conditioning to control for the effect of individual ancestry leaving out the locus under consideration (ANCOVA/IAE minus marlcer); and third using the complete individual ancestry estimate for the conditioning (ANCOVA/IAE). As shown in Table 2, eight of twenty-one (38%) of the markers showed significant differences (p < 0.05) among the three genotypes, including two of the four candidate gene markers (OCA2 and TYR). When using an alpha level of 0.05, only 5% of the marlcers tested were expected to yield significant results. As such, the fording of 38% significant difference indicates that population structure is related to both ancestry and pigmentation (Pfaff et al., sups°a, 2001, Parra et al., supra, 200I).

[0210] One way to remove the effects of population structure is to test for differences conditioning on the individual ancestry estimates (IAE). When the complete TAE
was used to condition (ANCOVA/IAE), only one locus showed signficant average differences among genotypes, OCA2, the human P gene. When a less conservative conditioning approach was talcen, in which the locus under consideration was left out of the individual ancestry estimate (ANCOVA/IAE minus marker), there were four significant results: OCA2, TYR, FY, and SGC30055.
[0211] A Bayesian full probability model for admixture and marker genotypes was also set up (McKeigue et al., sups°a, 2000). Score tests for linkage were based on testing for an independent association of pigmentation with number of alleles of European ancestry at each locus, one at a time, in a regression model that includes individual ancestry (estimated from marlcer data). The 1-sided probabilities for the score tests are shown in Table 2, where three loci showed evidence of linlcage to skin pigmentation at an alpha level of 0.05 f OCA2 (p = 0.005), AT3 (p = 0.027), and TYR (p = 0.033)}. To confirm these results, other markers informative for ancestry in OCA2 were identified and will be analyzed by the score test method. The concordance between these ANOVA results and the Bayesian admixture mapping results was encouraging, and both methods will benefit from the addition of new unlinked AIMs, which will increase the precision of the individual ancestry estimates.
[0212] The Spanish-American sample from the San Luis Valley CO, also was analyzed for linkage and association using the Bayesian and ANOVA methods (Table 3).
This analysis included 442 individuals who were typed for 1 S marker loci informative for ancestry (2 SNPs in the DRD2 gene treated as one locus). The CYP19E2 marker (located near MYOSA, a pigmentation candidate gene) showed strong evidence for linlcage with the ethnic difference in slcin pigmentation. However, this result should be interpreted with caution because, unless several closely linked markers informative for ancestry are used, the test for linkage is not robust to misspecification of ancestry-specific allele frequencies. SNPs around the MYOSA gene can be analyzed to confirm these preliminary results.

Testing for an effect of single-locus genotypes on pigmentation in an African-American sample Marker DELTA ANOVA ANCOVA ANCOVA Bayesian score AF vs IAE minus IAE test probability EU marker FY-null 0.997 0.000'' 0.004* 0.396 0.106 LPL 0.479 0.000* 0.130 0.501 0.167 WI-14867 0.448 0.061 0.561 0.253 0.100 AT3 0.575 0.001* 0.090 0.479 0.027*

Sb19.3 0.488 0.497 0.784 0.993 0.402 APOAl 0.505 0.507 0.718 0.317 0.488 D115429 0.429 0.578 0.833 0.442 0.139 RB1 0.611 0.743 0.535 0.128 0.102 OCA2 0.631 0.000* O.OOI* 0.019* 0.005*

NICIR 0.428 0.601 0.482 0.590 0.486 WI-16857 0.536 0.745 0.789 0.311 0.137 WI-11153 0.652 0.026* 0.225 0.698 0.223 SGC30055 0.457 0.001* 0.048* 0.291 0.061 WI-7423 0.476 0.193 0.544 0.169 0.058 WI-11392 0.444 0.032* 0.200 0.507 0.257 MID 154 0.444 0.260 0.439 0.728 0.250 Mm 187 0.370 0.121 0.302 0.296 0.450 DRD2 TAQ,D 0.553 0.226 0.273 0.247 0.471 GNB3 0.463 0.984 0.530 0.166 0.065 TYR-192 0.449 0.003* 0.014* 0.117 0.033*

GC 0.697 0.213 0.750 0.604 0.190 1 Marker indicates the Ancestry Informative marker used in the test. Marlcers shown in bold and italics are candidate genes for pigmentation (viz. OCA2, MC1R, TYR).
2 "DELTA" (b) is the allele frequency difference between African and European populations.
3 Analysis of variance significance level where sex is the only covaxiate.
4 Significance level for a one-way ANCOVA analysis using individual ancestry estimates (M) where the tested locus was excluded as the covariate.
Same as three except the M is based on all 21 markers.
6 Bayesian Admixture Mapping 1-sided probability.

Testing for an effect of single-locus genotypes on pigmentation in a Spanish-American sample Markers DELTAZ ANOVA ANCOVA ANCOVAS Bayesian score NA vs EU IAE minus markerIAE test p-value Mm-575 0.546 0.240 0.383 0.248 0.93 TSG1102055 0.744 0.027* 0.023* 0.366 0.39 Wl-11153 0.628 0.012* 0.036* 0.406 0.12 SGC-30610 0.427 0.093 0.133 0.565 0.13 WI-17163 0.521 0.192 0.268 0.967 0.17 WI-4019 0.296 0.875 0.762 0.319 0.93 WI-11909 0.663 0.026* 0.020* 0.146 0.13 TYR-192 0.417 0.006* 0.012* 0.129 0.07 DRI72Bcl 0.485 0.254 0.517 0.827 0.35 DRD2TaqD 0.582 0.418 0.461 0.550 0.35 D11S429 0.376 0.235 0.511 0.545 0.65 WT-14319 0.494 0.036* 0.054 0.061 0.30 CYP19E2 0.423 0.000* 0.000* 0.002* 0.001*

PV92 0.624 0.183 0.276 0.591 0.50 WI-7423 0.402 0.426 0.318 0.309 0.64 CKMM 0.545 0.257 0.569 0.579 0.55 ~ Marlcer indicates the Ancestry Informative marker used in the test. Markers shown in bold and italics are in or near candidate genes for pigmentation (viz. TYR-192 and CYP19E2 near MYOSA).
2 "DELTA" (8) is the allele frequency difference between Native American and European parental populations.
3 Analysis of variance significance level, where sex is the only covariate.
4 Significance level for a one-way ANCOVA analysis using individual ancestry estimates (M) where the tested locus was excluded as the covariate.
5 Same as three except the M is based on all 15 markers.
6 Bayesian Admixture Mapping 1-sided probability.

PAIRWISE POPULATION COMPARISON OF SNP DISTINCTION
[0213] Table 4 shows the results of genotyping and statistical analysis that demonstrate several different but important points (the sequences for each of the AIMS in Table 4 can be found by reference to Table 6, using the marker number.

Pair-wise population comparisons of SNP distinction AF- AF- AF- AF- AF- AF- CT- CT- CT- CT- CT- EA- EA- EA EA- SA- SA- SA- ME ME
PI-CT EA SA ME PI AI EA SA ME PI AI SA ME -PI AI ME PI AI -PI -AI AI marker 0.4 O.GV Q 5~ 0 4~~ 0.~ .0:3 0.3 0.2 0.1 0:3 0.1 0.1 0:2 0 0:4 0.1 0.2 0;3 0.3 0.2 0.4 959 ~~~ , 0.5, (~ ~,, ~ 5 0 ~ Ø4 0 0 0 0,3 0.2 0 0 03 0.2 0 0.3 0.1 0:3 0.1 0:4 0.1 0 0.1~~ 0.1 0.1 0.1 0.2 0 0 0.2 0 0.1 0.2 0.1 0.2 0 0.2 0: 0.2 0 0:2 9 62 0.3 0:4 0 2 0:4 0.3 0.1 0.1 0 0.1 O 0:4 0.1 0 0 05 0.2 0.1 0:3 0.1 05 0:4 963 09; 0.3f tl.~ , il 7~ 0.1 0 4~ 0 0.1 0 2~ 0, - 0.5 0.5 0.4 0 2 0.1 0 l tl~0 4~,. ~i<0.3 0:3 964 0.3 0 0.2 0.2 0 0.5 0:4 0:1 0.1 0.3 0.2 0.3 0.2 0 1 ~ 0 0:2 0.3 0.2 0.3 0:5 9 0.2 0.2 0.2 0.2 0.1 0.1 0 0 0 0 0.1 0 0 0 0.1 0 0 0.1 0 0.1 0.1 970 0.3 ",~t,'~;,(~*~ 0 4~,~ 13 0.4 0.2 0 0.2 0:2 0.2 0.3 O.I 0.1 0.2 0 0 0.2 0.2 0.5 0.2 0.3 0.5 0.1 _0.2 0.3 0.2 0 0.4 0.3 0.1 0.3 0.1 0 0.2 0.2 0.1 0.3 0,3 0 0.3~ 0:8'' 0.4 0 3~ p.7 0 ~ 0.4 0.1 0.1 0,3 0:3 0.3 0.5 0.1 0.2 0.2 0.2 0.2 0.4 0:3 0 973 0.1;- (i.~ 0.4 Oj 0.~ 0 4s 0 f~ 0.4 0 ~ Cr 0:5 0.2~ 0,0 0.1 0.4 0.2 0 1 .~t,i~;~ 0.5 0.1 975 0.4, D:~' 0.4 0 5;?0 # 0 ~~ 0.4 0 0.1 0.5 0.5 0.3 0.3 0.1 0.1 0.1 0:4 0.4 0.4 0:4 0 977 0 0.1 0.1 0 0.3 0.2 0.1 0 0 0,4 0.1 0.1 0.1 0:4 0.1 0 0:4 0.1 0.3 0.2 0.5 97 8 (10.3 0.2 0.4 0.1 0.2 0.3 0.3 0.2 0,5 0:4 0 0.1 0.2 0.1 0.1 0.2 0 0:3 0.2 0.1 ~~, 0,;6 0 '~'~ :,p,~ 0 5 p ~ ",t? 0.1 0.1 0.1 0 0.1 0.2 0.2 0.1 0.2 0 0.1 0 0.2 0.1 0.1 980 0.2 0.1 0.5 0 5 0.. 0:5 0.1 0.3 0.3 0,4 0:3 0.4 0.4 0:4 0:4 0 0 0 0 0 0.1 9 86 O.I O.I 0 0 0.2 0 0.2 0.1 0.1 0:3 0.1 0.1 0.1 0.1 0.1 0 0.2 0 0.2 0 0.2 993 0.1 0 0.1 0.1 0.1 0.1 0.1 0.2 0.1 0.2 0 0.1 O.I 0.1 O.I 0 0 0.2 O.I 0.1 0.2 0 0 0.1 0 0~2 0 0 0.1 0 0.2 0 0.1 0 0:2 0 0.1 0.2 0.1 0.2 0 0:2 1001 0 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0 0.1 0 0 0.1 0 0.1 0.1 0 0.2 0 0.1 1 003 0 0 0 0 0.1 0 0 0 0 0.1 0 0 0 0 0 0 0.1 0 0.1 0 0 1013 0 0 0 0 0 0.1 0 0 0 0.1 0 0 0 0.1 0 0 0.1 0 0.1 0 0.1 1015 0 0 0 0 0 0 0 0 0 0.1 0 0 0 0 0 0 0 0 0 0 0 1022 y0,.7 0 0 3j I? 7 0 5 O 1 0Ø4 0 0.2_ 0 $: 0.3~.._~?,~ 0:5 0.2 0.4 0.2 0.4 0 2t~~1 ,~~ 1029 .0>5 0 7 ~ 0 ~ lk ~ i? $~ tl.~, 0.2 0.1 0 0.3~ 0:3 0.1 0.2 0.1 0.2 0 1 0.2 0.3 0 3 0:3 0.1 1033 06 0 3~' 0 'T h:70.1~ 0 li 0.3 0.1 0.1 0.5 0 0.4 0.5 0.2 0.3 0 1~ Q~ 0 1~,y0 ~G,t 0.2 0:5 1034 0.2; 0.5 0.2 0 2 0.4w"(!,~' 0.4 0 0.1 0.2 0.5 0.4 0.3 0.2 O.I 0.1 0.2 0:5 O.I
0.4 0.3 1035 0.1 0.2 0 0 0.1 0.1 0.1 0.1 0.1 0 0 0.2 0.2 0.1 0.1 0.1 0.1 0.1 0.2 0.1 0 103 0.2 0 0.1 0 0 0 0.2 0.1 0.2 0.2 0.2 0.1 0 0 0 0.1 0.1 0.1 0 0 0 103 8 0.1 0.1 0.1 0 0.1 0.1 0.2 0.2 0 0 0 0 0.2 0.2 0.3 0.2 0.3 0.3 0.1 0.1 0 1 03 9 0 , 0 '~ 0.3 0.2 0.5 0 2~ U.'~ 0.3 0.2 0,5 0.2 0.4y 0 0 2( 0 0.1 0.2 0.1 0.3 0 0.3 1 040 ~ 0.6 _0_O.l;~t~.6 _0.2 0 3~ ti.'T 0.6 0.1 0,5 0.4 0.1, O,f 0.2 0.3 0.5 0.1 0.2 0.4 0.3 0.1 1041 0.5 ~0.7' tl.~ t? 6 0,~, 0.2 0.3 0.1 0.2 0.1 0:2 0.1 0.1 0 2 ~ 0 0 0.4 0 0.4 0.4 1042 0.3~, O.S 0.3 0 4 Qa~ 0 4 0.2 0 0 0.3 0.1 0.3 0.2 0 0.2 0.1 0:3 0.1 0.2 0 0.2 0.3' 080:3 03 0'x-00.5 0 0 0.4 0.2 0.5 0.5 0.1 0.3 0 0.4 0.2 0.4 0.2 0.2 1044 AF- AF- AF- AF- AF- AF- CT- CT- CT- CT- CT- EA- EA- EA EA- SA- SA- SA- ME ME
PI-CT EA SA ME PI A_I EA SA ME PI AI S_A ME -PI AI ME PI AI -PI -AI AI marker 0.3~ 0.~;' 0.1 0 3~ 0 6 x,0.3 0.3 0.1 0.3 0 2 ~G 0.4 0 0.1 0 2 X0.5 0:4 0.3 0.1 1047 0.2 p~~~ 0.2 0 1 0 0.1 0.4 0 0 0.4 0.1 0.3 0:4 0 0:5 0.1 0.3 0.1 0.4 0 0.5 1 ~ a ~Ua7 D ~ Q 9 ~ 9 ~ $ ~~~$ 0.1 0.2 0 2 0 1 0:3 0.1 0 1 0 0.2 0.1 0.1 0 0.1 0.1 0.2 1049 ~, (1.6 0,1~ (15 , l~ 7 0 0 3 0.5 0.1 0 1~ ~ 0:4 0.4~ ~ 0.1 0.1 0.2 0:5 0 3, ,~ 0:4 0.2 1050 0.3, ~.~~ 0.4 0 4 0 ~.TO 0.2 0 0.1 0:3 0.2 02 0.2 0 0 0 0.2 0.2 0.2 0.2 0 105 0.1 0.3 0:3 0.3 0 0,2 0.2 0.2 0:3 0 0.1 0 0.1 0:2 0.1 0.1 0:2 0.1 0:3 0.1 0.2 0.3 0.2 0.2 0:3 0.1 0 3 0.1 0 0 0 2y 0.X0.1 0.1 0.1 0:4 0 0.2 0.5 0 2~ t1,5 0.3 1053 A5 ### 0.3 0 4 0 3 0 3 0 ~~ 0.2 0.2 0,3 0.3 0.3 0.4 0.3 0.3 0.1 0.1 O.I 0.1 0.1 0 1055 0.4 0 'T 0.4 0 30 ~ ''II I~' 0.4 0 0.1 0.4 0.4 0.4 0:4 0 0 0.1 0:4 0.4 0:5 0.5 0.2 0 0.1 0.1 0 0 0.2 0.1 0 0.2 0.2 0,1 0.2 0 0 0 0.1 0.1 0.2 0.2 0 1057 0 0.1 0 0.1 0.1 0.1 0.1 0 0 0.1 0.1 0.1 0.1 0 0 0 0.1 0.1 0.1 0.1 0 1059 0.4 0.'~ 0 3~ Ii 5 ~ 0'~' 0;4 0.3 0.1 0.1 0;3 0 0.4 02 0 0.3 0.2 0.4 0.1 0.2 0.1 0:3 1060 ~:"~' OTC; tl ~, tl '7 0'~ EUØ1 0 0 0.1 0.1 0.1 0.1 0.2 0,2 0 0.1 0.2 0 0.1 0.1 1062 0.5 0.2 0.3 0 4 0 1~ ~ 0.3 0.2 0.1 0~4 0.3 0.1 0:2 0 lj x0.1 0.2 0.5 0 3 0 4~_tl,1064 ~., t~°fi (i ~ 0.4 0 4~ Qy$,_4Ø1 0.2 0.2 0,2 0.2 0;3 0.3 0.1 0.1 0 0:4 0.4 0.4.-0:3 0 1065 0.2 0.2 0.2 0.2 0.2 0.2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1066 0.1 0.3~'' , U~G ~i ~' 0.2 0 0.1 0.5 0:4 0.1 0.1 0:3 0.3 0,1 0.3 0 0.4~,(1,~:
0 4; ,I16 0.2 1070 0.3 0 0.1 0.2 0 0.1 0:3 0.2 0.1 0.3 04 0.1 0.2 0 0.1 0.1 0.1 0:2 0.2 0:9 0.1 0 0 0.4 0.5 0 0 0 0.4 0.5 0 0 0.4 0.5 0 0 0.1 0.4 0.4 0:5 0.5 0 1 072 0.1 0.1 0.1 0.1 0.1 0.1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1073 0.4 0 1 0 4~ ~,6 0 5'~1?*"l 0.3 0 0.2 0.1 0:2 0.3 0.4 0 4llw0.2 0.1 0:2 0.1 0.1 0.2 1074 { I17 r0 ~', '~'~_~~:0 8 ~t~0.1 0 0 0.1 0.1 0.1 0.1 0 0 0 0.1 0.1 0.1 0.1 0 0.5~ p~.~y-0 4p~~0.3 0.1 0.1 0.1 0 0.1 0.1 0.1 0 0.2 0.2 0.1 Ø3 0.1 0.1 0.2 0.3,, 0.T'' 0.4 03~p$_tl0.4 0 0 0.4 0:4 0.4 0.4 0 0 0.1 0:4 0.4 0:5 0.4 0 1077 0.3 0.1 0.1 ### 0.3 0.1 0.2 0.3 0.3 0.1 0.2 0.1 0.1 0.1 0 0.1 0.2 0 0:3 0.1 0.2 1078 0.3 (~ 6 0.2 0 Z~ 0.~ 0.4 0.3 0.1 0.1 0.2 0.1 0.4 0.4 0.1 0.2 0 0:3 0.2 0:3 0.2 0.1 IO80 0.3 ' O:~T 0.4 0 3~ 0 7 t1 ~ 0.5 0.2 0.1 0.5 0.3 0.3 0:4 0 0.1 0.1 0.3 0.2 0.4 0:3 0.1 1081 0.3 0 0.3 0 2 O.~a 0.4 0.2 0.1 0.1 0.2 0.1 0.3 0:3 0 0.1 0.1 0:3 0.1 0.3 0.2 0.1 1082 0.5' 08..fl~ 031,.D?w'0~ 0.4 0.2 0.2 0:2 0.4 0.2 0Ø1 0.1 0.3 0.1 0.3 04~
00.2 1083 0 05 01 01 03 0.2 0.5 0.1 0 0,4 0:2 0:4',. ~l0.1 0.3 0.2 0.3 0.1 0.4 0:3 0.2 0~ ~ 0 7'~Il ~ i~ $ 137 ~~,'~ 0.2 0.2 0 2 0 2 0.2 0.1 0 0 0 0.1 0 0 1 0 0 0 P ~
Q~ 0 1~,;' t# 6 4:7 0 1 0 4~ 1~ ~ 0.2 0I# f' 0.4 0.4i 1?.C0 0.2 0.1 0.4 0 2~~~~Ø4 0.2 1087 ~ .. ~'s'~ 0.1 0 5# ~,~ 0 2'11 (1,~~,, 0.2 0.2 0.5 0.1 0.4 0.4 0.1 0:5 0 0.3 0.1- 0.3 0.1 0.4 1088 "DELTA" (~) values for select AIMS from the list claimed are shown.
The AIM unique identifier is shown in the last column.
Cells having numbers in bold (not shaded) indicate good b values; cells that are shaded with numbers in bold represent extremely high ~ values.
AF-African, CT-European, EA- East Asian, SA-South Asian, ME-middle eastern, PI-Pacific Islander, AI-Native American.

[0214] First, Table 4 was derived from screening several hundred candidate AIMs electronically selected from the public databases, thus demonstrating that only a minority of the candidate AIMS from the public databases are real AIMS. As discussed above, the public SNP databases were electronically screened to find good candidate AIMs (since frequency data is provided for three "racial" groups, though the level of admixture for these groups is not known). Second, 384 individuals of various continental and BGA origins were genotyped at each of these sites: 70 African samples collected from Nigeria and Congo, 65 European samples collected from Northern Europe, 70 East Asian samples collected from recent immigrants to San Francisco, CA; 35 Middle Eastern samples collected from Turlcey, 35 South Asian samples collected from India and 25 Pacific Islander samples collected from the Philippines and US Samoa).
[0215] A sampling of the data for about 70 AIMs that passed the screening process from Z 75 candidate AIMs screened is shown in Table 4. The delta (8) value is a measure of how well the sequence for a polymorphism enables one to predict membership to one or the other group; i.e., how distinct the two populations axe with regard to the sequence at this polymorphism. ~ values are shown for 69 of the 175 AIMS; the other 105 had S
values of 0 for each pair-wise population comparison and, therefore, were not tt~.ie AIMS.
AIM 1068 in Table 4 is representative of types of failure (zeros across all pair wise comparisons - some AIMS with zeros across the population pairs are present in Table 4 because they are informative for populations not shown in this particular table). This result confirms that most of the candidate AIMs culled from the public database are not true AIMs and highlights the value of the present invention.
[0216] A relatively large investment in genotyping and analysis required to identify which candidate AIMs are true AIMS. While this process can be bypassed, e.g., one could simply genotype samples at 100 candidate AIMs and extract the data from those that prove to really be AIMS, genotyping is an expensive procedure and, as such, the waste incurred would make the test economically impractical. To develop an economical and practical test for ancestral proportions, the test must query large numbers of true AIMS. Many publicly available candidate AIMS are not true AIMS due, fox example, to low sample sizes, such that allele frequencies in the public databases are simply not very reliable, though they do offer some information. While it is expected that the frequency of true (i.e., validated) and well-characterized (i.e., population specific frequencies known with certainty) AIMs in a random collection of SNPs would be about 5%, the frequency of true AIMS in a culled set of candidate AIMS is about 50% and, after proceeding as disclosed herein, the frequency of true AIMS in a collection of SNPs is 100%
[0217] Second, the results of Table 4 demonstrate that some of the AIMs are good for the resolution of Africans vs. Europeans, other AIMS are good for the resolution of Native Americans vs. Africans, etc. Though selected based on European/African/Asian allele frequency differentials, some AIMs provide good distinction of other groups such as Pacific Islanders, South Asians and Middle Easterners. This-type of information can only be learned by genotyping in larger samples, and a test for ancestral proportions must go through this step in order to be accurate (for example, if the test only worlced in the 3-dimensions allowed by data that is publicly available - European, African and Asian - the results obtained, for example, for a Hispanic would be ambiguous). The panel of SNPs in Table 4 provides a well balanced mix of AIMS with resolution power for each of the possible pair-wise comparisons for 7 population groups, and this panel would constitute a good test for ancestry proportions.
Data for south Asians, Middle Easterners and Pacific Islanders do not exist in the public databases and, therefore, was generated for these studies. In comparison, one attempting to develop a test for ancestral proportions in 7-dimensions by simply sequencing at candidate AIMS haphazardly selected from the public SNP databases (i.e., without selection through data production) would need to compile a battery of thousands of SNPs to obtain a panel such as that in Table 4 because certain pairs of populations are difficult to resolve (e.g., South Asians and Europeans, which constitute a larger IndoEuropean group united by a common language base).
[0218] The results obtained using the panel of AIMs shown in Table 4 is shown in Table 5. The presently disclosed algorithm (see Example 6, Table 12) was used to calculate the proportions for a group of 96 individuals who reside in the southeastern United States, and who describe themselves to be Caucasian.

Ancestral proportions for a large number of self reported Caucasians oho oho % M F MGM MGF PGM PGF CO
SELF

EUR97 NAM 0' ' ca ca ca ca ca ca ca us EUR89 EAS 1.1 0; ca ca ca ca ca ca ca us NAM

EUR100, O EAS 0 ca ca ca cc ca ca ca us NAM

EUR76 NAM 24 0 ca ca ca ca ca ca ca us EAS

EUR80 NAM 20 0 ca ca ca ca ca ca ca us EAS

EUR1OOEAS 0'NAM 0 ca ca ca ca ca ca ca us EUR92. OAFR 8: ca ca ca ca ca ca ca us NAM

EUR95 EAS 5 NAM ;0 ca ca ca ca ca ca ca brazil EUR88 NAM 12 0 ca ca ca ca ca ca ca us ,> EAS ~

EUR.100 0 EAS 0 ca ca ca ca ca ai_caca us NAM

EUR76AFR 24 0 ca ca ca ca ca ai ca us NAM ai ai EUR100 Q EAS 0 ca ca ca ca ca ca ca us NAM

EUR100: p EAS 0 ca ca ca ca ca ca ca us NAM

EUR.51 4$~EAS1 ca ca ca ca ca ca ca us NAM;, ~

EUR100: EAS . ca ca ca ca ca ca ca us EUR100: OEAS 0 ca ca ca ca ca ca ca us NAM

EUR-100 0 NAM 0 ca ca ca ca ca ca ca us EAS

EUR54 NAM 45 0 ca ca ca ca ca ca ca us ' EAS

EUR9~ NAM 5' ' ca ca ca ca ca ca ca us EUR'1 QO 0 EAS . ca ca ca ca ca ai_caca us EUR89 NAM 11 0' ca ca ca ca ca ca ca us EAS

EUR100 - O 0 ca came ca ai_caai_meNULLNULL us NAM EAS

EUR88'NAM 12 0 ca ca ca ca ca ca ca us EAS

EUR100 0 EAS 0 ai_caca ai_caca ca ai ca us NAM

EUR100 O NAM 0 ca ca ca ca ca ca ca us EAS

EUR76NAM 24 0 ca ca ca ca ca ca ca us EAS

EUR75 NAM 14 11 ca hi ca hi hi ca ca us ~ EAS

EUR100 O EAS 0 ca ca NULL ca ca ai ai us NAM
~

EUR67 NAM 33 0 ca ca ca ai_caca ca ca us EAS

EUR72 NAM 28 0 ca ca ca ca ca ca ca us EAS

EUR73 NAM 27 0 hi hi hi hi hi hi hi us EAS

EUR77 NAM 23 0 ca ca ca ca ca ca ca us EAS

EUR66AFR 34 0 NULLai_caca ca ai-caca ca us NAM

EUR99'NAM 1 EAS . ca ca ca ca ca ca ca us EUR67 NAM 33 0 ca ca ca ca ca ai_caca us EAS

EUR100 O NAM 0 ca ca ca ca ca ca ca us EAS

EUR100 0 EAS 0 ca ca ca ca ca ca ca us NAM

o/a. % % M F MGM MGF PGM PGF CO
~ SELF

puerto EUR 88..NAM12EAS0 ca ca ca ca ca ca ca rico EUR 1.00 O 0 ca ca ai ca ca ai_caca us NAM EAS ca E 90,EAS 10, 0 ca U ca ca ca ca ca ~
R '- NAM ca us EU 78 NAM 22 0, ca ca ca ca ca ca ai_caus R _ EAS

EUR;100 OfEAS0 ca ca ca ca ca ca ca us NAM

EUR 76 NAM 14 10.ca ca ca ca ca ca ca us EAS
~

EUR 76 NAM 24;EAS0 ca ca ca ca ca ca ca us EUR 100 . 0 ca ca ca ca ca ca ca us , NAM

EUR 9T NAM 3 Q' ca ca ca ca ca ca ca us EAS

EUR 91NAM ' . ca NULL NULL NULL NULL NULLNULL us .

EUR 94 NAM ~ 4 ca ca ca ca ai ca ca us EAS

EUR 100 0 0 ca ca ca ca ca ca ca us NAM EAS
~

EUR 1 OO OEAS ' ca ca ca ca ca ca ca us NAM ~ 0 EUR 97 EAS 3'; 0 ca ca ca ca ca ca ca us NAM
' EUR 100 0 0 ca ca ca ca ca ca ca us NAM EAS
' EUR 100 O 0' ai_caai_caai ai ai_caNULLNULL us EAS NAM

EUR 100.NAMO 0 ca ca ca ca ca ca ca us EAS

EUR 60 NAM 40 0 ca ca ca_hica ca hi hi us EAS

EUR 84AFR 16EAS0 ca ca ca ca ca ca ca us EUR 100 O D ca ca ca ca ca ca ca us NAM EAS

EUR 99'NAM PEAS 0 ca ca ca ca ca ca ca us EUR 89 NAM 11 0 ca ca ca ca ca ca ca us EAS

EUR 67 NAM 32 '1 me me me me me me me us J EAS

1NAM55 EUF~40 ~ ca ca ca ea ca ca ca us EMS

EUR 84EAS OAFR 16 ca ca ca ca ca ca ca us EUR 74 NAM '13 13 ca ca ca ca ca NULLNULL us EAS

EU 88NAM 12 : ca ca ca ca ca ca ca us R EAS 0.
' EUR'100 0 0 ca ca ca ca ca ca ca us NAM EAS

EUR 98 NAM 0'EAS2 ca ca ca ca ca ca ca us ' EUR 100 0 0 ca ca ca ca ca ca ca us NAM EAS

EUR 70 NAM 30 0 ca ca ca ca ca me ca us EAS

E 86: 14' 0 ca ca ca ca ca ca ca us U EAS NAM
R

EUR . 65: 16EAS19 ca ca ca ai_caca ca ca us NAM

EUR 100 0'EAS0 ca ca ca ca ca ca ca us NAM

EUR 100:EASp 0 ca ca ca ca ca ca ca us NAM
' EUR 85 NAM 15 0 ai_caai_caai_caai ca ca ca us EAS

EUR 100'NAM0'EAS0 NULLca ai_caca ca NULLNULL us EUR 72:NAM 28 0 ca ca ca ca ca ca ai us EAS

AFR 89 EUR 11 0 aa_caas as as aa_caas aa_caus EAS ca EUR 84 EAS 10 6 ca ca ca ca ca ca ca us NAM

EUR 95 NAM 5 0 ca ca ca ca ca ca ca us EAS

EUR.98. 2EAS 0 ca ca ca ca ca ca ca us NAM ~

EUR 97 EAS 3 0 ca ca ca ca ca ca ca us NAM

EUR 62 EAS 19 19 ca hi ca hi hi ca ca us NAM hi %.. %' % SELF M F MGM MGF PGM PGF CO

puerfio EUR70EAS 12NAM 18 hi hi hi hi hi hi hi rico EUR74 NAM 26 0 ca ca ca ca ca ca ca us EAS

EUR94 NAM 0'AFR 6 ca ca ca ca ca ca ca NULL

puerto EUR65EAS 11 24 hi hi hi hi ca hi hi rico AFR

EUR75 NAM 25 0 ca ca ca ca ca ca ca us EAS

EUR91 NAM 9 EAS 0 ca ca ca ca ca ca ca us EUR1 OO ' O 0 ca ca ca ca ca ca ca us NAM EAS

EUR7.1 29 0 ai ai_caca ca ai ca ca us NAM EAS ca ca EU 72 NAM 28 R ca ca ca ca ca ca us EAS 0 ca EAS56 NAM 44 b ca ca ca ca ca ca ca us EUR

The percentage of ancestry is given after each group: EUR - IndoEuropean, NAM -Native American, EAS - East Asian/Pacific Islander, AFR - African. The self reported race for each individual (SELF), their mother (M), father (F), maternal granchnother (MGM), maternal grandfather (MGF), paternal grandmother (PGM), paternal grandfather (PGF) and the country of their birth is shown.
[0219] The results of Table 5 demonstrate that most people who describe themselves to be Caucasian are, indeed, of majority IndoEuropean ancestry as determined using the BGA test, i.e., using the panel of markers in Table 4 and the algorithm (Example 6, Table 12). About 40% of these Caucasians were measured as 100% European ancestry, with no admixture, whexeas 60% showed detectable admixture. By way of critical evaluation, the proportions obtained can be compared with self reported admixture. For individuals of Table 5 that claim all of their parents and grandparents are unmixed Caucasians ("ca" across all columns), the rate at which 90% or greater European ancestry is found is higher (55%) for persons who report no mixture in their pedigree than for persons who report mixture in their pedigree (35%). The fact that half of those reported no mixture in their pedigree illustrates how little most people know about their BGA, at least in anthropological as compared to geopolitical terms.
[0220] The public databases employed a small number of samples for each of the three groups (African-Americans, Europeans and Asians). Thus, from the public databases, the actual allele frequencies for the claimed SNPs is uncertain and was only determined with accuracy from the present work. Further, the use of African-Americans as a parental group is faulty since, as disclosed herein, they are an admixed population (between Africans and Europeans). The best way to find SNP markers that are useful for the present methods is to genotype a number of samples from the maj or BGA groups of the world for all of those of apparently different minor allele frequency between at least two of the groups, calculate the 8 values, and rank them.
FAMILY PEDIGREE
[0221] The BGA method for the use of SNPs (AIMS) to determine ancestral proportions was applied to the examination of a family pedigree of a father, mother and their 3 children (see Figures 6 and 7). The man was mostly European and his wife is Mexican, so, if the test is accurate, their three children should plot somewhere between the man and his wife and this is exactly what was observed. Three of the father's grandparents on the father's side of the family are relatively pure Greek/European, but one was almost pure Cherokee.
All four of his grandparents on his mother's side were European mix. The plots for the mother (Figure 6B) and father (Figure 6A) are shown. By drawing lines from the vertices that bisect the triangle edge opposite that vertex, where the vertex represents 100% and the triangle edge 0%, it can be seen that the father is about 85% European, 11% Native American, 4% African (he was of no detectable East Asian, South Asian or Pacific Islander ancestry). The 11%
Native American seems to have come from his paternal grandparent. The percent that would be predicted, knowing that 7/8 of his great-grandparents were European mix and 1/8 were Native American, is 12% (1/8), which is in good agreement with the data produced from the test. The mother is from Mexico and is of Hispanic descent. As discussed previously, Hispanics are an admixed population with European and Native American contribution. It can be seen here that the mother is of 11% European, 76% Native American and 13% African descent (there was no detectable East Asian, South Asian or Pacific Islander ancestry).
[0222] Because the three children each received one chromosome from each of their mother and father, they should each plot somewhere in between him and his wife. Using the method the children plot as expected (Figure 7). It is clear from these results that the point estimate of the children is between that of the parents, as would be expected.
From these results it seems clear that Child #1 (Figure 7A; $0% European, 1$% Native American, 2%
African) has ancestral proportions more similar to his father than his mother, whereas Child #2 (Figure 7B; 61 % European, 31 % Native American, 8% African) and Child #3 (Figure 7C;
54% European, 37% Native American, 9% African) are of ancestral proportions about half way between their two parents.
[0223] Although each child receives a chromosome from the mother and one from the father, the children are different because the mother has chromosomes of mostly Native American descent, though some have European and African flavor as well, and the father has chromosomes of mostly European flavor though some are of Native American flavor as well.
When a child is conceived, he or she receive a chromosome from each parent, but which of the two the child receives from the mother is random (i.e., "independent assortment"). Since some of the mother's chromosome pairs have a member of the pair with European flavor, some of the children will receive the "European flavored chromosome" and other children will not. From the present study, it is clear that Child #1 (Figure 7A) received more of lvs mother's European flavored chromosomes than did the other two children. Thus, while each child is a 50/50 blend of maternal and paternal chromosomes, their ancestral proportions are unique and random functions of those of their parents.

FOUR WAY ADMIXTURE ESTIMATE OF ANCESTRY
[0224] This example demonstrates that a four way admixture BGA test provides the same results obtained using three 3 way BGA tests.
[0225] As indicated above, BioGeographical Ancestry (BGA) is the heritable component of race. Because socio-cultural and geo-political metrics for measuring human race are hmnan, not natural, constructs, their use in genetics research makes it difficult to control for population genetic structure, and may obscure important correlations between BGA and human biology. This example provides methods and compositions to accurately measure genetic structure within individuals. The human genome was mined for candidate Ancestry Informative Markers (AIMs), which were validated on an ultra-high throughput genotyping platform and used to establish parental population allele frequencies. Using 71 of the most informative AIMS (Table 6), which cover most of the chromosomes, and coalescing the human population to four main continental population groups (sub-Saharan African, East Asian, IndoEuropean and Native American), the MLE method was used to determine individual BGA admixture proportions and their associated confidence intervals. As disclosed herein, self reported population affiliations correlated almost perfectly with the majority BGA population affiliation determined for a sample of 2,024 international samples.
BGA admixture results were surprisingly frequent, and when observed, were generally not inconsistent with anthropological and geopolitical history. The admixture proportions produced tracked in family pedigrees in a manner consistent with the law of independent assortment, and simulation revealed that the markers relevant for resolving the group affiliations functioned independently within the confines of our algorithm.
Because a large number of high 8 value markers were used, the test was surprisingly robust;
reasonable levels of simulated allele frequency errors that could be caused by biased parental sampling had no significant impact on the BGA proportions determined. These results demonstrate that BGA
admixture can be reliably deternlined from a DNA sample.
Sample Collection:
[0226] Parental Samples - For establishing the parental group allele frequencies, 100 relatively homogeneous descendents of four crudely defined human population groups were genotyped. These four groups corresponded to a coalescence of a simplified human pedigree back in time to a point where populations were relatively isolated to the sub-Saharan African (sub-Saharan Africans), Europe and the Middle Eastern (IndoEuropeans), North/South American (Native Americans) and East Asian (East Asians) continental regions.
In essence, the extremes of the existing populations has been taken in terms of physical characteristics and known human migration patterns, and the human pedigree simplified by assuming that, because the rest of the population exhibits physical features along a continuum defined by these extremes, all of humanity arose from radiation within and admixture between these four main continental groups.
[0227] Collection efforts were focused on individuals residing in each region and of relatively homogeneous affiliation with descendents of each group in terms of self described "race"; each subject exhibited a strong physical appearance associated with descendents of each group and reported homogeneous affiliation with that group. There exists no de facto final arbiter of BGA affiliation, or race, but collecting without regard to ethnicity might introduce systematic errors in frequency estimation for a given group if systematic admixture is a function of ethnicity. Where possible, an attempt was made to collect from as wide a variety of etluucities as possible within each parental group, expecting admixture within each parental sample to balance out. Though it would be better to collect from individuals of known homogeneous affiliation, if sampling was not biased in terms of admixture or other population structure, the admixture extant to samples that are practical to collect from would tend to reduce the power of the test rather than introduce systematic bias.
The existence of Hardy-Weinberg Equilibrium for each marker within each BGA group was relied on as an indication that a reasonably good sample was obtained. The sub-Saharan African samples were collected in Nigeria and Congo, Africa; the European samples were collected from various locales in the United States; the East Asian samples were collected from Japan and China; and the Native American samples were collected from "Nativos"
inhabiting a remote region of southern Mexico. All samples were collected under IRB guidelines for the purposes of genetic studies of human population variation.
[0228] Experimental Samples - After reading and signing an approved IRB
consent form, subjects completed a biographical questionnaire and provided either a buccal swab or 4 mI of blood. On the questionnaire, the subjects described themselves, their mother, father and maternal and paternal grandparents as belonging to the "African", "American Indian", "Asian", "Caucasian", "Hispanic" or "Other" group, with the option of reporting "Don't Know" for each family member. For some of the subjects, digital photographs were taken;
explicit permission was received from those subj ects whose photographs were presented.
DNA was extracted from circulating lymphocytes or buccal swabs using commercial bits (Qiagen), and a primer extension protocol employing a 25K SNPstream ultra-high throughput (UHT) genotyping system was used (Orchid Biosciences).

Estimating BioGeographical Ancestry (BGA) [0229] A software program was written based on the algorithm of Hanis et al.
(sups a, 1986) for using multilocus AIM genotypes to determine the Maximum Likelihood Estimate (MLE) of individual BGA admixture (Example 6; see, also, Table 12). The delta (8) value is an expression of the ancestry informativeness of the marker (Dean et al., sup~~a, 1994). For a biallelic marker, the frequency differential (~) is equal to pX - py, which is equal to qy - qX, where pX and py are the frequencies of one allele in populations X and Y and qX and qy are the frequencies of the other. To test the departures from independence in allelic state within and between loci, the MLD exact test was used (Zaylcin et al., Gev~etica 96:169-78, 1995).
[0230] The collection of 71 AIMS used for in this Example was selected to maximize the cumulative 8 value within, and miumize differences in the cmnulative 8 value between each of the six possible pairs of the four dimensional (sub-Saharan African, Native American, IndoEuropean and East Asian) problem. The algorithm inverts the population specific allele frequencies to obtain a likelihood estimate of proportional affiliation corresponding to a multilocus genotype using three groups at a time (mainly for computational convenience and because a 4-dimensional admixture is likely to be relatively rare). For example, the likelihood of 100% IndoEuropean, 0% Native American, 0% East Asian is calculated, then the likelihood of 99% IndoEuropean, 1 % Native American, 0% East Asian is calculated next, and so on until all possible IndoEuropean, Native American and East Asian proportions are considered, then the process is repeated for all possible IndoEuropean, Native American and African proportions, and all possible Native American, African and East Asian proportions.
The likelihood of maximum value is selected as the Maximum Likelihood Estimate (MLE).
[0231] When plotting a single MLE on a triangle plot, the space within which the likelihood is within 2-fold, 5-fold and 10-fold of the MLE is delimited (these intervals are not plotted when multiple MLEs are shown in a single triangle plot). For calculating the MLE
using all four of the BGA groups together, the procedure was executed in the same exact manner; instead of 3 possible 3-way BGA combinations, only one 4-way BGA
combination is possible. All of the MLEs described in this Example were calculated using the 3-way calculation scheme. Alternative versions of this type of test are possible, for example, using different AIMS and different parental groups corresponding to different coalescences of the human pedigree, which would be expected to provide results meaningful with respect to a different anthropological time scale than those provided here.
[0232] At the time the database was screened, the SNP Consortium (TSC) had contributed data on approximately 27,000 SNPs where frequencies were available on three populations (African American, European American, and East Asian). This database was screened for candidate AIMS, i.e., SNPs of 8 >0.40 between any two of the four continental population groups (see Example l; see, also, Shriver et al., sups°a, 1997).
Parental samples of sub-Saharan Africans (AA), IndoEuropeans (IE), East Asians (EA) and Native Americans (NA) were screened for each of the 200 candidate AIMs with the largest b values and, of these, 71 were validated as true AIMS (i.e., true SNPs), with minor allele frequency greater than 1%
and of b > 0.40 for at least one of the group pairs. The 71 AIMS are shown as SEQ ID
NOS:l to 71; the top 100 candidate AIMS for the group pairs were as follows:
EA by AA
(SEQ ID NOS: SEQ ID NOS:7, 2I, 23, 27, 45, 54, 59, 63, and 72 to 152); EA by IE (SEQ ID
NOS:3, 8, 9, 1 l, 12, 33, 40, 59, 63, and 153 to 239); and IE by AA (SEQ ID
NOS:1, 8, 1 l, 21, 24, 40, 172, and 240 to 331). It should be noted that some AIMS identified by one pairwise comparison also can be AIMS for a second pairwise comparison (e.g., SEQ ID
N0:59 was identified as an AIM for EA by AA and EA by IU comparisons), although such AIMS are an exception. In addition, many of the 7I AIMs are not in the list of the top 100 candidate AIMS shown for any of the pairs (but were in the top 200 candidate AIMs);
candidate AIMs were not used, for example, because they did not genotype well due to the SNP type of amplification parameters used for the exemplified platform, or for other reasons as disclosed herein.

Pair-wise 8 values for the 71 AIMs used in the BGA test AF- AF- AF- CT- CT-CT EA AI AI EA-AIAIM SEQ ID NO:
EA

0.330.2140:24 0.120.09 0.03 958 31 0:T510.2070.64 0.540.11 0.44:960 65 0.5640.5490.53'0.010.04 0.02 961 52 0.2860:3640.08 0.080:36 ~ 963 39 ~
, 0.44 -.

0.8450:86 0.36 0:560.48f0.0g 964 70 . .

s . ' 0.-327' 0.01 0.14'' '0.47-966 37 0.464 0.33 ' ~

0.3210.09T0:49 p.42~0. 0.58 969 63 ~ 17 ' 0.1630.1710.06 0.010.11 0.11 970 40 Y

0.3-3107 0.57 0:37' 0.13 971 64 ' 0.24' 0.497;0.1860.16 0,310.34 0.03 972 1 0.3390.7790,6~ Q:440.~8 0.16 973 22 0:3720.7250:85 0.350:48'0.13 977 57 ' ~

0.0410.1220.18 0.080.14 0.06 978 3 0.549.0,268'0.19 0:280.36 0.07 979 34 0.6060.6710.52 0.070.08 0.15 980 21 0.1020,7'71:0.610.670.51'0.16 993 54 0.6230.0710.09 0.550.53;0.02 1000 18 ~

0.0380.1 0.14 0.060.1 0.04 1015 24 ...

0.6310.2650.06 0 0-57 p.2' 1022 20 0,6740.0250.14 0:65-0g2 x;17 1029 13 0.5.?30.7140:87 0.190.35 0.16 1033 41 0.634Q.2950:6 0.34;0.03 0.31 1034 46 ~

0.1720.5390:63 0,37'0.460.09 1035 25 4E-04'0.7110.17'0.710.17 0.54'1040 29 ' 0.650.0310:29 0.680:36 : 1041 S
0.32 0.340.535:0.38 0:220.06 O.I6 1043 35 0.3340.8070.52 0.4fi0.19 0.29:1044 16 i ~
~

0.3430.6860.55 0.340.21 0.1 I 047 61 0.1680.5370.08'0.370.08 ' 1048 19 0.45=
A

0.6860.8 0:96-0.11Q.27~0.16 1049 26 0.606'0.1140.25 0.490.35 0.14 1050 48 '' 0.5860.58 0.240.~3 0.01 1051 42 0.348 0.2630.171'0.26 0.090.52 0.43 1053 62 0.5340 0.27 0.530.26 0.27 1055 67 0.3610.7420.76 0.380.4 0.01 1056 44 0.1530.0220.03 0.180.18 0.01 1057 67 0.0430.0430.04 0 0 0 1058 59 AF- AF- AF- CT- CT
CT EA AI AI EA-AIAIM SEQ ID NO:
EA

0.3920.7250.39 0.330' 0.33 1060 27 0.7290.8290.59 0.1 0.14 0.24 1062 23 -p.4560.193'0.78 0.260.32 0:58 1064 10 ... .
.

0.1570.1570.16 0 0 0 1066 7 0.0430.0430.04 0 0 0 1068 11 0 0.0130.11 0:340'.440.1 1071 53 326 ~ ~

0.0710.0710.07 0 0 0 1073 8 0.'7Q40.8090.81'0.1 0.1 0 1075 69 ' 0.5350.34 0.070.13 0,2 1076 71 0.44 0.3480.728v0.74 0.3$0:39 0.01 1077 45 0:261'0.7360.6x 0.:47Ø350.13 1081 43 0.3.430;5360.4110.190.06 0.13 1082 36 ~

0.4540.8210.:8'70<370.42 0.05 1083 6 0.036Ø45Ø17 Q.49- 0:28 1084 51 0.21 0.74 0.117'. 0.620.39 0.?3 1087 28 0.35 _ 0.9851 Ø99;0.010.01 0.01 1111 30 . ..

0.8510:2520.3.7Or6 Ø4 0.12 976 68 0.4950.0930.15 0:590.64;0.06 1113 4 0.1330:2 0;67 0:070.54 0:47 1116 50 0.58 0.40'70.49 0.140:05 0.09 1117 47 0.5670.650.64 0.080.07 0.01 1120 60 0.36$0:3790.47 0.750'.84'0.09 1121 33 0.5;'750;6640:59 0.090.02 0.07 1122 58 ~

0.1770.0860.3 0:2,6~ 0:21 1124 17 0.48 0.41:60.0890.02 0:330.39 0.07 1128 2 .

0.1080::4490:51 0.560.62 0.07 1036 9 ' ~

0.1490.1 0 0:25,0:25 0 1130 55 .1;

0.4730.6340.61 0.160.14 0.02 1136 38 0.8920.667Q.22 0:230:68 0.45'1137 56 0:0760.2790.69 0.~ 0:61 0:41 1138 66 0.6I20.660.62 0.050.01 0.04 1139 32 0:4350.6 0.29 0.170.14 0:31 1140 ~

0.4870.5640.61 0.080.12 0.04 1141 49 0.1550.5710.31 0:730:46 0.26 1146 15 AF- sub-Saharan African, CT - Indo European, EA - East Asian, AI - Native American.
The AIM unique identifier is shown (AIM) as well as the GENBANK accession number provided by the authors upon submission of the SNP sequence to the NCBI:dbSNP
database.
The number of AIMS with 8>0.40 for each pair-wise comparison is shown at the bottom of the list.

[0233] The 71 AIMS used in the exemplified panel were spread throughout 21 of the 23 autosomal chromosomes (Figure 8), with the average chromosome containing 3 AIMs (see Table 6). Each had alleles in Hardy-Weinberg equilibrium, both overall with all four BGA groups considered together, and within each BGA group, and none were found to be in linkage disequilibrium with one another. The software program used individual genotypes for these AIMS with a maximum likelihood algoritlun (see Example 6, Table 12;
see, also, Example 1). The use of the 71 markers with this algorithm provides another example of the "BGA test".
[0234] The BGA test was used to calculate BGA admixture proportions for the parental Native American, African and Indo European samples used in the construction of the test.
After calculating the admixture proportions for each sample, they were plotted in a triangle plot to allow for the relative proportions of a 3-way mixture to be represented in two dimensions. Because these were the same samples that comprised the parental groups and from which the population allele frequencies were derived, they were expected to exhibit relatively homogeneous BGA (i.e., of low admixture) and, in fact, the sub-Saharan Africans, Native Americans and European parental samples all registered with relatively homogeneous BGA (i.e., they plotted towards the appropriate vertices of a BGA triangle).
[0235] The BGA test was next used to determine BGA proportions for 1,186 individuals of self reported race (43 African Americans, 1,120 Caucasians, and 23 Hispanics). 306 of the individuals (26%) showed homogeneous BGA (100% for any one group). 101 of the 1,186 (8.5%) harbored >5% BGA affiliation for three groups, indicating that, for these individuals, a modification of the software to perform 4-group calculations might be more appropriate, and the vast majority of the samples were chaxacterized by 2-way admixture.
Significantly greater European admixture was identified in African Americans relative to that of Nigerians (visualized as a dispersal of points away from the sub-Saharan African vertex). In contrast, the IndoEuropean samples plotted as neatly in the IndoEuropean vertex as had the parental samples, though low levels of Native American or East Asian admixture were not uncommon - roughly two-thirds of subjects harbored detectible, though generally low levels of such admixture. The Hispanic subjects plotted in an even distribution along the Native American/IndoEuropean axis, consistent with the l~nowledge Hispanics arose from the blending of colonial Europeans and resident Native Americans about 500 years ago.
[0236] In order to determine whether, and to what extent, the major proportions obtained with the BGA test corroborated with self reported race, BGA admixture proportions were calculated for 2,048 individuals of self reported race, and blindly (in a computational sense) compared the majority BGA determined from the test against each individual's self reported majority race. A very strong concordance was observed between the major BGA
group determined with the test and the self reported majority race (Table 7). Using the test, 1252/1252 self described European-Americans (U.S. born Caucasians) registered with majority IndoEuropean BGA. 191 of 201 self described African Americans showed majority sub-Saharan BGA, with the remaining 11 showing sub-Saharan BGA as the minor affiliation with IndoEuropean as the majority affiliation. Hispanics showed roughly equal distribution in majority BGA between IndoEuropean and Native American, consistent with the results observed in the triangle plot and the anthropological history of this group.

Comparison of Majority BioGeographical Ancestry with Self Reported Race Self Re oi-tecl"Race' a ~urope~nAfricanNative AmericanEast Asian Euro ean-American US white1252 0 0 0 Africans from Ni eria 0 217 0 0 African-American US black 11 190 0 0 Natives "Nativos" from 1* 0 72 0 Guerrero, Mexico His anics born in US, of 101** 0 90 0 Mexican herita a Chinese 0 0 0 10 Ja anese 0 0 0 10 Other eneral Asian from 1 *** 0 0 33 US

South Asians from India 36 0 0 0 *Minor proportion was Native American * * Significant Native American as second type.
* * *Minor proportion was East Asian [0237] Even when unexpected results were obtained, such as the finding that one individual from Southern Mexico was majority European, the results were more or less concordant in that the expected affiliation was the minor affiliation, rather than being absent altogether, and the major affiliation (Indo European) made sense in light of the history of the region (colonized by the Spanish hundreds of years ago). In this particular case, the level of Native American ancestry was only slightly less than 50%. Not one gross error was observed where, for example, a self reported European-American was classified as majority East Asian. Though the 11 self described African-Americans of majority IndoEuropean BGA
would appear to be a gross error, the results were similarly concordant - each of these African American samples exhibited roughly equal IndoEuropean/African proportions, suggesting admixture, consistent with the historical tapestry of the region from which the samples were talcen, rather.than test error.
[0238] To conduct a truly blind test (as opposed to a test that is blind in a computational sense), the San Diego Police Department Crime Lab (SDPD) and the National Center for Forensic Science at the University of Central Florida (UCF) each submitted ten buccal swabs of numerically encoded identity for BGA tests. The BGA test was performed and the results returned to SDPD and UCF, each of which independently evaluated their results and revealed the self reported population affiliation of the sample. The major percentages determined from the BGA admixture proportions test were not inconsistent with the self reported population affiliations (see Table 8). Several of the samples were from individuals affiliated with groups that are logically considered to be admixed - e.g., Filipinos (SDPD2, SDPD3, Table 8), African American or Carribean (SDPDS, SDPD6, UCF7, UCFB, Table 8), Mexican Americans (Hispanic, SDPDB, SDPD10, Table 8) and Puerto Rican (UCF6, Table 8);
significant admixture was detected for each of these samples. Moreover, the type of admixture detected was reasonable with respect to the anthropological history of the affiliated population. For example, people of sub-Saharan African, Native American and IndoEuropean descent live in Puerto Rico, whereas East Asians are relatively rare, and the test results for the Puerto Rican individual tested showed IndoEuropean and Native American, not East Asian admixture.

[0239] In addition to determining the MLE of admixture, the software program was designed to survey the probability space, and define that space within which the likelihood of proportional affiliation was 2-fold, 5-fold, and 10-fold less likely to be the correct answer than the MLE (confidence contours, which are plotted on the triangle plot as rings around the MLE). One way to test the accuracy of the maximum likelihood algorithm for determining MLE and confidence contours was to observe whether and how these values change when certain of the AIM markers are eliminated from the analysis by replacing the genotypes for each with "failure" readings. For example, if for a given sample genotype, all of the genotypes for rnarlcers of high ~ value for the African/East Asian distinction were replaced with "failuxe" or "no data", an accurate test would be expected to show warped confidence contours only in this dimension of the triangle plot. Accordingly, the BGA
test was used to plot one sample of majority East Asian BGA with its associated confidence intervals (Figure 9A), then all 24 of the markers in the test were eliminated with informative ~ for Native American vs. East Asian BGA, and the MLE and confidence estimates were recalculated with the remaining AIMs (Figure 9B). Upon recalculation with the missing AIMS, the confidence rings were dramatically skewed from the East Asian towards the direction of the Native American BGA vertex (Figure 9B), indicating, as expected, that the laclc of AIMS with good ~ values for East Asian/Native American distinction produced an estimate for which the confidence along the East Asian/Native American axis was not high.
Presumably because the sample typed as of majoxity East Asian, and ATMs for the distinction of East Asian/IndoEuropean and East Asian/African affiliation were left unmolested, the MLE shift itself was minimal; most of the uncertainty along the East Asian/Native American axis was apparent in the shift of the contours. Similar experiments with other samples and AIMS produced similar results.

Blind Challenge of BGA Test by the San Diego Police Department and the Center for Forensic Science at the University of Central Florida ID# % Admix Ratio % % Self Rennrt~prl Racy SDPD1 EUROPEAN _96EAST-ASIAN p NATIVE-AMERICAN4 Euro ean-American SDPD2 EAST-ASIAN 53 EUROPEAN 47NATIVE-AMERICAN0 Flli ino SDPD3 EAST-ASIAN ' NATIVE-AMERICAN28EUROPEAN 11Fili ino SDPD4 EUROPEAN 100EAST-ASIAN d NATIVE-AMERICAN0 Euro ean-American .

SDPD5 AFRICAN 69 EUROPEAN 31NATIVE-AMERICAN0 Caribbean '' SDPD6 AFRICAN 67''EUROPEAN 33EAST-ASIAN Q African American SDPD7 EUROPEAN 99 EAST-ASIAN 1>.NATIVE-AMERICAN0 Euro ean-American .

SDPD8 NATIVE-AMERICAN57 EUROPEAN '43EAST-ASIAN 0'Mexican American SDPD9 EAST-ASIAN 86 NATIVE-AMERICANQ.EUROPEAN 1'4Fili ino SDPD10NATIVE-AMERICAN38 EAST-ASIAN 28EUROPEAN 36~Mexican American UCF1 EUROPEAN 90.EAST-ASIAN _0NATIVE-AMERICAN1t?.'Ukraine/Ital UCF2 EAST-ASIAN 98.NATIVE-AMERICAN2 EUROPEAN 0.Chinese UCF3 EUROPEAN ,:88EAST-ASIAN ' NATIVE-AMERICAN9 Ukraine 3 a UCF4 EUROPEAN 1.D0EAST-ASIAN 0 NATIVE-AMERICAN0 Euro ean-American UCF5 EUROPEAN 87 EAST-ASIAN Q NATIVE-AMERICAN13:Greek UCF6 EUROPEAN 62 NATIVE-AMERICAN38EAST-ASIAN p Puerto Rican .

UCF7 AFRICAN 83 EUROPEAN 17EAST-ASIAN 0 African American UCF8 EUROPEAN ".5gAFRICAN 31'NATIVE-AMERICANQ''Jamaican UCF9 EUROPEAN 84 EAST-ASIAN 12NATIVE-AMERICAN4-Finland UCF10 EUROPEAN 98 EAST-ASIAN 2 NATIVE-AMERICAN0:,''Scotland :' [0240] In order to determine the reproducibility and consistency of the BGA
admixture proportion determinations, five samples were genotyped and analyzed on separate occasions.
One sample was chosen of self reported majority affiliation with the European American, African American, Hispaiuc and Asian groups, and a fifth sample was chosen from the parental Native American group. With the exception of failed loci, the genotypes at each marl~er in each individual were 100% consistent between runs, indicating that the ATMs genotype reliably. A 1-3% variation in BGA admixture proportions was observed from run to run; simulation studies showed that the variation was attributable to these genotyping failures. This result indicates that the failed loci did not pose a significant barrier to reproducibility of the BGA admixture determination for an individual, either in terms of majority BGA or admixture levels. From these simulations, it was determined that the BGA

test tolerates about ten locus failures if the samples are of least binary admixture, and for samples of no admixture, larger numbers of failed loci are tolerable (i.e., change in admixture percentages were less than 5%).
[0241] In order to determine if the BGA admixture proportions determined with the test were sensible given the rules of family inheritance, proportions for three generations from several family pedigrees were calculated. A typical result is shown in Figure 10, which depicts the ratios obtained for a family of confirmed paternity (using STR
tests) with substantial European/Native American admixture. The first generation individuals were self reported European-Americans that were determined with the BGA test to harbor significant Native American admixture, which was passed to their son and daughter in different pxoportions that were not inconsistent with the law of independent assortment. The son's spouse was a Hispanic native of Mexico, and was determined to be of 26%
European/74% Native American admixture. Each of their offspring harbor roughly intermediate levels of Native American and IndoEuropean admixture between their parents, again not inconsistent with the law of independent assortment. One of the sons typed as of a small percentage of East Asian ancestry, but the level (4%) was close to the reliable limit established as discussed above (about 3%). Other pedigrees tested (n=8) showed similarly concordant results, indicating that, in terms of majority BGA and admixture levels, the BGA
test results were sensible within the context of family pedigrees.
[0242] Because the BGA test relies on parental population allele frequencies, the extent to which the admixture proportions produced by the test were influenced by parental allele sampling bias was investigated. A pair of populations was selected, as were the AIMS
relevant for resolving affiliation between these two populations (Table 6 -selected those with the highest 8 values for this pair of groups), then the allele frequency was adjusted for each of these AIMS in one of the groups such that the 8 value for the AIM was reduced by 20% (with respect to these two groups). In effect, the power of the test was intentionally degraded by 20% for the resolution of affiliation between a specific pair of groups; this degraded test is referred to as the Ancestry 2.lEA/EU BGA test, where the EA/EU refers to degradation in the distinction between the East Asian (EA) and IndoEuropean (EU) groups. 31 samples were randomly selected, the genotypes were run against the Ancestry 2.lEA/EU
BGA test in exactly the same manner as for the original (ANCESTRYbyDNA~ 2.0) test, and the xesults were compared with those obtained with the ANCESTRYbyDNATM 2.0 test.
[0243] If the results from the ANCESTRYbyDNATM 2.0 test were highly sensitive to parental allele frequency error caused by sampling bias, which might be expected to be on the order of a few percent at the most, the 20% change introduced for the Ancestry 2. lEA/EU
AIMS should result in admixture proportions substantially different from those of ANCESTRYbyDNA~ 2Ø Since the number of IndoEuropean/East Asian mixes observed was significantly greater than other types of mixes, such as African/Asian mixes or Native American/African mixes, the European/East Asian pair was selected for the first test -Ancestry 2.lEA/EU - and the BGA group pair for which the number of AIMS and cumulative ~ value is the lowest is Native American/East Asian, the 8 values for this pair was altered in the second test - Ancestry 2.1NA/EA. The average change observed in admixture proportions between ANCESTRYbyDNATM 2.0 and Ancestry 2.lEA/EU was I.4%
(Standard Deviation 2.44%). For the Native American/East Asian pair, the average change between ANCESTRYbyDNATM 2.0 and Ancestry 2.1NA/EA was I% (Standard Deviation 2.3%).
[0244] Socio-cultural or self held notions of race are not likely to be as tightly linked to human biology as BGA, which is the heritable component of race. Because it is subjective, imprecise and sometimes inaccurate, the use of self identified race for the inference of BGA, as is currently practiced, obscures how and why human biology is related to human anthropology. Furtlzermore, the rigid binning of patients into prefabricated racial groups is a practice of generalization that is wholly unsatisfying because many individuals can trace their origins to multiple populations through the process of admixture. A
repeatable, testable anthropological approach to define BGA can provide a means to draw connections between BGA and heritable diseases, whether through straight correlation and/or better study designs, ox more subtle means such as through gene mapping methods that rely on the admixture process, such as MALD.

[0245] As disclosed herein, a 7I marker test allowed a determination of BGA
proportions and their confidence intervals. The test enabled a determination of the relative proportionality of BGA within individuals, thus distinguishing the BGA test from other tests previously used for inferring ancestry from DNA. In terms of the majority BGA
affiliation, more than 2200 tests were performed and no result was obtained that was inconsistent with self held notions of race. Previous tests of ancestry have been accurate only to the upper 90% range (Shriver et al., supf a, 1997; see, also, Frudalcis et al., supra, 2003, which is incorporated herein by reference). The enhanced performance observed with the BGA test can be because CODIS and other STRs that have been commonly used for infeiTing ancestry from DNA were not selected for their 8 values, but were selected for their polymorphic complexity in the world population. For the BGA test disclosed herein, the entire genome was systematically scanned and the best AIMS for this purpose were selected.
In addition, most efforts to infer ancestry from DNA using STRs or Alu sequences have attempted to categorize or bin samples into single "racial" groups. For individuals of extensive admixture, such as a 50/50 mix, such a method would seem to produce a "wrong" answer as many times as a "right" answer. In contrast, with the BGA test, ancestry is determined in terms of proportional affiliation, thus ameliorating this problem.
[0246] The BGA test is distinguishable from other tests in that it employs SNPs that cover most of the chromosomes. Pan-chromosomal coverage using the BGA test provides a substantial advantage over tests using CODIS STRs, which only cover a fraction of the chromosomes. In addition, the BGA method appears to be the first that quantifies the confidence limits fox its answers. The BGA test as exemplified is largely heuristic, and divides the world into four main anthropological groups that fall largely along continental lines. Though the geographical divisions are respectful of the anthropological history of human migrations, the use of four groups is indeed a simplification of a very complex situation and can be considered to be arbitrary. Further, the problem of determining proportional aff liation has been simplified by calculating the most likely 3-way (rather than 4-way) combination, because individuals of 4-dimensional BGA are thought to be rare, and because it is more convenient in a computational sense. However, while more complex tests may be able to capture more of the extant detail of anthropological history, even a crude 4 population test provides data of meaningful and historical content, provided the results are interpreted strictly with respect to these divisions and to the parental samples used in the construction of the test.
[0247] The choice of specific parental groups and of dividing the world into certain groups simply provides a point in coalescent time by which to scale the inferences provided by the test. In fact, the difference between the answer provided by a test based on 4 world population groups and a more complex test based on 25 would be one of anthropological time scale, not "accuracy". For example, most Hispanics born in the United States that have been tested and most individuals claiming American Indian heritage type that were tested revealed minor Native American admixture on an IndoEuropean background. However, unlike the case for Hispanics, some of the individuals claiming American Indian heritage were typed as of minor East Asian admixture instead of Native American admixture. Because the founders for Native Americans migrated from East Asia, possibly in different waves at different times in history, the genetic distance between Native Americans and East Asians is lower than between Native Americans and sub-Saharan Africans or IndoEuropeans (Cavalli-Sforza and Cavalli-Sforza, supra, 1995). Among pre-colonial North Americans, proportional affiliation to East Asian or Native Americans would be expected to be different for individuals whose ancestors were part of the first waves than from those whose ancestors were part of later waves.
[0248] The parental samples for Native Americans used in the present study were derived from southern Mexico, and the AIM allele frequencies established for Native Americans obtained from this sample might be expected to be more representative of the ancestors from earlier waves of migration across the Strait. Native Americans from Latin America and South America would likely be more closely affiliated with early wave ancestors than, e.g., those from North America such as the Aleut Indians (and others), which would probably be more closely affiliated with ancestors from later waves across the Strait. The East Asian affiliation for those individuals claiming affiliation with American Indians may be a by-product of the choice of Southern Mexico Nativos as the parental source, and the use of only a 4 group anthropological scheme, but nonetheless, the answer is not a "wrong"
answer in a scientif c sense. Rather, it reports affiliations with respect to a coalescent time scale defined by the source of the parental samples and the anthropologically meaningful way in which the world was divided for this study.
[0249] A different test, with more markers to resolve affiliation within the North American groups from each other and from East Asians, would opexate on a different coalescent time scale and likely classify these individuals as of minor "American Indian" or "North American Native American" achnixture. Nonetheless, the fact that the exemplified BGA test revealed that a significant number of individuals of "Native American" ancestry show more affiliation with East Asians than Native Americans, as defined by the metrics built into the test, may be yet another example that social or human history based notions of population affiliation are semantic, subjective and not always accurate in a biologically meaningful way. Even though Aleuts may appear to resemble East Asians as much as, or even more than, most Native Americans in terms of physical features, and even though they are indigenous to a geographical locale as proximal to East Asia as to temperate North America, they are considered by most to be North American Indians and by extension, Native Americans because their home lies east of the Bering Strait. Similar examples have been observed for certain other population groups, as discussed below, illustrating the disconnect between the measurement of population affiliation using genetic markers and that from geographical and social borders man has devised to ascribe racial identity.
[0250] With deference to the qualifications of the BGA test, it is interesting to compile the results in order to extract some meaningful anthropological and/or sociological knowledge.
Using the BGA test, 11 of 201 African Americans tested showed major IndoEuropean and minor African BGA, and Puerto Ricans of majority African descent almost always describe themselves as Hispanic, again pointing towards the deficiency with current notions of race as a dichotomous entity based on man-made constructs. As was observed in the present study, Risch et al. (sups°a, 2002) and Rosenberg et al. (supf°a, 2002) have shown that, when tested against methods of reporting BGA that rely on genome markers, majority population affiliation is quite accurately reported on questionnaires. The present testing of over 2,000 individuals indicates that the majority BGA affiliation can be accurately predicted from the self reported race, that discordance between the two is not a very significant event, and that determination of majority ancestry affiliation is not the main problem with current self reporting methods. However, perhaps the most surprising result was the extent of admixture for each population tested. When individuals claimed admixture, it was almost always confirmed with the BGA test; each case of expected sub-Saharan African and East Asian admixture was confrmed with the BGA test, and every Hispanic of Mexican descent registered with either major or minor Native American admixture, as expected.
[0251] Slightly more than two-thirds of all "Caucasians" tested exhibited minor East Asian or Native American admixture, and virtually none of these individuals reported any significant pedigree admixture on his or her questionnaire. Some of this admixture appears to be a function of ethnicity. Not only did individuals of relatively homogeneous self reported Northern and Eastern European heritage more commonly show East Asian BGA, but Rosenberg et al. (supf°a, 2002) showed there to be significant structure within the "European-American" or "European" population that support this observation; specifically they showed that Russians commonly exhibit minor East Asian heritage. There are multiple times in history where such East-Asian/IndoEuropean admixture may have taken root, including, for example, the Mongolian invasions of Europe, and the Caucasoid expansion to Scandinavia, whose Lapp inhabitants were derived from Northern Asians, exhibit Mongolian features, and share a common history and culture with East Asians (Cavalli-Sforza and Cavalli-Sforza, supra, 1995). Some of the Native American admixture observed is concordant with history;
that Filipinos exhibit extensive Native American admixture, for example, is not entirely surprising in view of the Spanish having conquered most of Latin America and exported Native American slaves to these islands, which were a Spanish territory until recently. The extent of Native American admixture observed for some Filipinos was quite high, perhaps reflecting that Native American admixture is relatively common in this part of the world, and that the pedigree for many Filipinos is dominated by large numbers of individuals of relatively low Native American admixture, rather than recent admixture with individuals of r highly polarized BGA proportions. The Native. American admixture commonly observed in "Caucasians" likely came from a blending of European and Native American peoples in North America.

[0252] In most cases of systematic admixture, such as between Scandinavian/Russian IndoEuropeans and East Asians, sub-Saharan African/IndoEuropean in the U.S., Native AmericanlEast Asian mixture in the Philippines or IndoEuropean/Native American in the U.S., the geographical proximity and/or historical mixture of the relevant groups is well established by history. For instance, an African/East Asian mixture was rarely observed in these studies, and history knows few examples of times when these two populations lived in close proximity to or mixed with one another. The type of admixture observed was also interesting when compared to self held notions of "race". For example, African Americans were more admixed with IndoEuropean Ancestry than were Caucasians with African ancestry, highlighting a difference in how Caucasians and Africans view their heritage, and involving recollections of the "one drop rule". Given the extent of admixture observed, it is lilvely that hidden or cryptic BGA structure arising from the process of admixture (a process that is not completely documented and quantifiable from our anthropological literature, as it is based on human constructions) is of potential concern for creating gross (or finer) structure differences between groups of study samples. Such a difference in structure would be expected to reduce the efficacy and power of large population based study designs.
[0253] Given that the majority ancestry was not inconsistent with self held notions of affiliation for over 2,200 blind test subjects of polarized (low admixture) and self repoz-ted race, the question arises as to how to assure the accuracy of minor admixture proportions such as these, and how such accuracy can be measured given that there is no ultimate arbiter of BGA in existence (except, possibly, genealogical information, see below).
Several experiments were performed that address this question and, when considered together, indicate that the minor proportions are accurately determined. First, the minor admixture percentages transmit along family pedigrees in a manner consistent with the genetic law of independent assortment. By definition, a large, unbiased error would malve it such that minor proportions were incoherent in the context of family pedigrees, i.e., the results would not be possible given the law of independent assortment, assuming the proportions of the parents to be correct.

[0254] Second, the minor admixture proportions are consistent, on average, with self held notions of admixture. If there was a large, systematic and unbiased error in the estimation of BGA affiliation, this error would impact the integrity of the minor proportion percentages more than that of the major proportion percentages, since most individuals are of relatively polarized BGA affiliation, but the correlation of minor admixture proportions/presence with self held notions of admixture would probably be very weak, if not imperceptible altogether.
This was not observed to be the case. For example, if the error rate was as high as 20%, as many individuals reporting minority Hispanic ancestry would show minor sub-Saharan African and East Asian ancestry as Native American; the present results clearly demonstrate this is not the case.
[0255] Third, of about 2,200 samples blindly tested from North America, an individual with majority East Asian with substantial (>10%) sub-Saharan African admixture, or vice versa, has never been observed. This non-observation is relevant because such individuals axe exceedingly rare in North America, from which the validation sample was derived. If there was a large, unbiased and systematic error rate, East Asian/African mixes would be observed as frequently as European-Asian mixes, which were frequently observed.
[0256] Fourth, a highly significant correlation was observed between minority IndoEuropean admixture and slcin melanin content in African Americans (see Example 1). If there was a large, unbiased error rate, such a correlation would not lilcely be obtained.
[0257] Fifth, when the true allele frequencies for AIMS relevant for the resolution of affiliation between two groups is adjusted, such that the parental 8 value was decreased by 20% for each relevant AIM for a given pair of groups, the overall power of the BGA test was degraded with respect to resolving affiliation proportions between these two groups, but essentially the same results were produced, both in terms of major affiliation and, more importantly, of minor admixture estimations. Parental sampling bias can cause inaccuracies in parental allele frequency estimation and 8 values in this manner, though certainly less than 20% given that the parental sample for the present studies was comprised of about 100 individuals. Further, the error in allele frequency estimation would have to be obtained in the same direction for most of the AIMS relevant for resolving affiliation between a pair of groups for such an error to exist. Nonetheless, this result indicates that, even if such an error existed, the performance of the BGA test Would be relatively unaffected. In other words, this experiment demonstrated that the BGA test is relatively robust in the face of parental sampling bias and allele frequency estimation.
[0258] Sixth, the distortion of confidence contours along only the axis coiTesponding to that for which genotyping failures were relevant shows a healthy disconnect among the interdigitating components (subsets of AIMS) of the test. In other words, if samples are fit to several templates for the determination of proportional fit, the elements of these templates should be independent for the test to be valid, and tlus was the result obtained.
[0259] It is difficult to imagine how significant systematic and unbiased error could exist in light of the above six observations. While each observation, on its own, may not prove the point of accuracy, taken together the results provide ample evidence that there is Little or no systematic and unbiased error in the disclosed BGA test . It nevertheless could be argued that the error in the test is not random, but biased in a linear manner. However, the fifth observation (above) argues against this possibility. For example, the finding of minor East Asian admixture on an IndoEuropean background was more frequent than had been expected, and such a result could occur if there was the "right" amount of allele frequency estimation error in the "right" number of marlcers in the "right" direction (an unlikely, but not impossible, situation). However, because the AIM markers used for these studies are not mutually exclusive to particular group pairs, such an error would manifest itself in the resolution of affiliation for many pairs, not just one, and the first four observations (above) indicate that this is not the case. Also, such an error would seem to require parental sampling from a highly inbred group. While great care was taken to avoid such sampling, every element of population structure was not controlled for because there exists no test to do so a p~io~°i. As such, a simulation was performed to estimate the contribution of linear error to the test results (fifth observation, above). The fact that admixture results were relatively impervious to substantial (20%) reductions in ~ value demonstrates that the quantity and quality of the AIMS used in the present studies were of adequate power such that reasonable levels of sampling error that may have been expected did not have a significant detrimental impact on the quality of the results. In terms of marker quality, this result is plausible because, given the selection process for AIMS, some of the best markers in the genome were used to determine BGA affiliation. In terms of marker quantity, these results are consistent with the observations that the results from the 71 marker test, as disclosed herein, were very similar to that produced from an earlier 30 marker test (Shriver et aL, supra, 2003). Thus, reducing the number of markers, while retaining the average marker quality, did not impair the test. Furthermore, in another study, reducing the quality of the makers, but not the quantity, produced the same results. Overall, these observations indicate that the BGA
proportions produced by the test are accurate with respect to the confidence intervals presented, and that the test performs in a robust manner.
[0260] The present results demonstrate that BGA admixture is more common than previously believed. If true, then it should be asked whether finer levels of BGA admixture axe linlced to human biology, e.g., drug responsiveness or disease predisposition. Such "cryptic" structure can only be determined using a molecular test because, unlike crude population structure. a questionnaire cannot be used. Finer levels of population structure beyond the more crude continental levels measured using questionnaires are recognized. For example, significant anecdotal evidence suggests that redheads require a 20%
larger dose of many common anesthetics, and exhibit a tendency towards hypertension and bleeding while under anesthesia (Cohen, The Seientist 16:10, 2002). These complex physiological responses would be difficult to explain based on melanocortin-1 (MC1R) variants previously linked to some of the variation in red hair color (Robbins et al., Cell 72:827-834, 1993; Smith et al., J. luvest. De~matol. 111:119-122, 1998; Flanagan et al., Hum. Molec. Genet.
9:2531-2537, 2000). There are likely specific gene variants responsible for these clinical phenotypes and, if these variants correlate with elements of population substructure or microstructure, as they appear to do, any study attempting to identify linkages or LD between markers and the relevant phenotypically active loci will be challenged from the outset at the study design step.
The exemplified BGA test can be extended using AIMS in addition to those disclosed herein to quantify the elements of population structure relevant for this particular problem because a precision and objectivity greater than that provided by self reporting of socio-cultural race will be needed to identify the elements of structure that can interfere with the design of genetics experiments.

APPLICATION OF THE BGA TEST TO GENEALOGY
[0261] This Example demonstrates that BGA admixture estimates can be integrated with genealogical information obtained using traditional genealogical research methods.
[0262] Genealogists collect data that largely is relevant in a geopolitical context (e.g., data relating to which countries a person's ancestors are fiom, what their religions were, and their last names) rather than in an anthropological context (e.g., what type of population admixture characterizes the person's family tree). There are two main sources for obtaining minority ad~.nixture in one's results: 1) recent exogamous admixture events; and 2) ancient affiliations with ethnic groups that are characterized by systematic admixture.
[0263] The results of an exogamous event is determined in a recent genealogical time (e.g., the last 250 years). For example, as shown in Figure 11, a Chinese great grandparent in an otherwise homogeneous IndoEuropean family tree would produce a grandchild of IndoEuropean/East Asian admixture. The individuals that are 100% East Asian (Chinese) are shown with shading (Figure 11), and the admixture results for the male (square) at the bottom of the pedigree (short arrow) axe-of interest. A person with a single 100%
East Asian great grandparent and seven 100% IndoEuropean great grandparents would be expected to have 12.5% East Asian admixture. By the law of genetic assortment, the expected level is actually a range around 12.5%, with values several percent above and below possible.
The grandparent indicated by the long arrow is about a 50%/50% East Asian/IndoEuropean mix and her daughter, the subject's mother, is expected to be a 25%/75% East Asian/IndoEuropean mix (Figure 11).
[0264] A~.zcient affiliations (i.e., considered with respect to an anthropological time frame) have been preserved in modern times by endogamous, relatively geographically isolated, close l~nit community structure (i.e., ethnicity). For example, modern day demography is shaped by not only the migrations our ancestors made to establish new populations, but also by admixture between these populations throughout the world. The map below shows these migrations as measured from Y chromosome sequences, which occurred over many tens of thousands of years.
[0265] Admixture between the groups, after each had developed as distinct groups, has occurred many times throughout ancient and even more recent history, and was represented by arrows on the map representing migration patterns. Fox example, there was extensive East Asian admixture in Russians and Eastern Europeans (Rosenberg et al., supra, 2003), and the extent to which the Mongolian and Hun invasions may have contributed to this admixture over a long time period remain a mystery. There was an even more pronounced East Asian admixture apparent for Native Americans (Rosenberg et al., supf°a, 2003); arrows were not included on the map fox this admixture because far too many would have been required and most of the admixture events are not know. Nonetheless, a person with a fair number of Native Americans or Russians in their family tree could very well exhibit as much East Asian admixture as an individual with a 100% Chinese grandmother and three other 100%
IndoEuxopean grandparents.
[0266] Although not shown, a time scale was constructed showing the time the most significant migrations occmTed, and was correlated to a very, very large family tree. The tree is for a single individual, who resides at the bottom apex of a triangle graph; it is large because it goes back 60,000 years when there are tens of thousands of ancestors for this person. The time scale for the migrations applies to the large family tree as well. The tree was the same as that shown in the pedigree map (Figure 11), only much larger and without the lines connecting the ancestors (spots represented each ancestor, but there were so many that it was not practical to show all of the lines connecting the spots. A
pool of spots represented "Russian", which for purposes of this example were assumed to be an ethnicity that arose about 18,000 years ago. Additional spots represented East Asian, and was based on the assumption that the average Russian harbors 10% East Asian admixture. A
third set of spots represented the precursors to Russians; these precursors are unknown but, for purposes of this example, were assumed to be Eastern Europeans.

[0267] In this example, most of this person's Russian ethnicity came from the left side of the family tree, which can be assumed to be that part of the tree representative of the subject's father's side of the family. If, as this example indicated, the average Russian harbors I O%
East Asian admixture, and half of the person's family tree is predominantly Russian, the person would be expected to harbor 5% East Asian admixture. East Asian admixture is signif cant for this person even though neither the person's grandmother, grandfather, or any other relative within the past 18,000 years was homogeneous East Asian. The way to visualize this on the family tree is to count all of the "East Asian" spots and divide them by the total number of spots in the tree, to arrive at about 5%. Thus, relatively homogeneous East Asians represented about 5% of the total number of ancestors for this person. Of course, the family tree for some people involves numerous groups that axe characterized by small degrees of this type of admixture. Family trees lilce that exemplified axe polarized with certain ethnicities, and it is uncommon to see a tree with an equal distribution of each of the four BGA groups (sub-Saharan African, Native American, IndoEuropean and East Asian) because, until recently, and even now to a certain extent, people have tended to have children with others like themselves. As such, most family trees are not a "mish-mash"
of random affiliations, but are highly polarized as exemplified.
[0268] The Pennsylvania Dutch provide another example of admixture due to ancient affiliations, wherein, in this community, it appears that there existed significant East Asian content prior to 1700 in German antecedents. These antecedents established communities that populated the valleys in the upper Rhine River Basin, then, later, more further inland.
Since these communities remained relatively isolated, the level of East Asian admixture has remained around the 20% level. Dilution of this level would require external admixture with other IndoEuropean ethnicities such as French or Sardinian, for which East Asian admixture is not detectible. In this case, the German antecedents were of substantial average East Asian admixture, perhaps due to sampling from within a more heterogeneous German population.
[0269] Most genealogists are interested in the type of admixture found in source arrived at according to the methods of exogamous admixture, wluch gives information about geopolitical affiliation of recent ancestors, rather than anthropological information on distant ones. This is because there exists little paper data for more distant ancestors compared to recent ones; the further baclc in time a person goes, the larger number of ancestors exist, mal~ing research difficult if not impossible; and the contribution of distant ancestors towards a modern persons genetic constitution is less on average per relative than that of more recent ancestors. As such, genealogists tend to seelc information such as that which might be produced due to recent admixture. For example, if a person is trying to prove or disprove a rumor or legend of American Indian ancestry, a 10% Native American admixture result would be very useful if it could be assured that the mechanism for this admixture was due to recent exogamous admixture events; and not to ancient affiliations with ethnic groups that are characterized by systematic admixture. While the BGA test does not allow a distinction between exogamous v. ancient admixture, the results of the test provide an important piece of the genealogical puzzle.
[0270] For some genealogists, depending on the family tree, evidence may strongly suggest that the mechanism of admixture is from recent events. As such, for a person whose family has paper evidence of an American Indian great grandparent, a 10%
Native American admixture result according to the BGA test can indicates the event likely arose due to recent admixture. In comparison, for a person of conf nned and homogeneous European ancestry, a 10% Native American admixture indicates the event likely arose due to an ancient admixture.
[0271] Negative results carry different meaning than positive ones for genealogists. For example, if there is circumstantial, but low quality, data suggesting a pure blood African great-grandfather, and the BGA test reveals 100% IndoEuropean, then the rumor would be discounted (taking into account the genetic law of independent assortment, which would make such a result possible, but unlikely if the data was in fact correct).
However, if a person's family is suspected to have had a Chinese great grandfather, one cannot prove it from a 20% East Asian admixture result, since it is not possible to distinguish exogamous admixture from ancient admixture.
[0272] It is important for a genealogist to bring other knowledge to bear in order to reconstruct the most likely source for an admixture result. In fact, the BGA
admixture data serves as an independent clue for one attempting to reconstruct a family history, and when it is used with genealogical knowledge, the two combine to form evidence that is more powerful than either on their own. As such, the BGA test provides an ancillary tool that can help fashion a system tailored for the genealogy community by providing BGA
admixture results in a manner that places equal emphasis on the anthropological sources, which transmitted ancient or very old (relative to a genealogical time frame, which encompasses the last 250-300 years) admixture to us in modern times, and exogamous admixture due to events in the family tree in the last 200 years.
[0273] A database of several thousand BGA profiles is built from people of various locations throughout the world such that one can query the database with a profile, plus or minus a pre-selected error range. A list of places for which this type of profile has been commonly found can be provided, or, for example, a map of the world that is color-coded can be provided, wherein the colors indicate the likely regional affiliations corresponding to the admixture profile/range. In other words, given a BGA admixture profile, a map can be provided showing the places from which the person's recent ancestors could have been derived. A person with 10% East Asian and 90% Indo European would show high probability of ancestral derivation from China (exogamous admixture) or Russia (more ancient admixture coupled with ethnic homogeneity over the family tree).
[0274] Similarly, the genealogist can provide a map, similarly color-coded, that is derived from paper research, which is based on geopolitical rather than anthropological information.
The two maps can then be overlaid, and Bayesian statistical calculations made that combine the information from the BGA test with that of the paper genealogy to provide a most likely estimate of recent family history.
[0275] By way of example, a person with 90% IndoEuropean and 10% East Asian BGA
and a paper genealogy of Romanian/BritishlSpanish ancestors would first query the database with his or hex BGA result, and be provided a map where the sources would be shown as possibly from East Asia (due to recent admixture), Russia and Northern/Eastern Europe (both due to a laxge number of more distant ancestors from relatively isolated and admixed groups).
The color-coding would give the probability of derivation from the regions based on the frequency with which the compatible BGA groups are found in each region, and it may be quite complex depending on the mixture type and the character of our database, which is a function of world-wide sampling. Second, the person would provide (or be provided with) a separate map based on the probable Romanian/British/ Spanish heritage documented from genealogical research, using map drawing tools we could provide. From this map, it would be apparent that the likelihood of recent, homogeneous East Asian ancestors is not high.
Third, a program would determine that the most likely origin of the 10% East Asian admixture is from the Romanian ancestors (not the British or Spanish, and not due to a Chinese grandparent, for example).
[0276] This type of presentation allow a person to learn the most likely source of an unexpected admixture result, using prior knowledge obtained through other means such as genealogical xesearch. This is valuable to a genealogist seeking to explain the derivation of a genetic constitution. Without this ancillary tool, a person with 90%
IndoEuropean and 10%
East Asian admixture would have no ready means to determine whether the test suggested a recent Chinese or Japanese grandparent/great grand parent or a certain ethnic aff Nation for which minor East Asian admixture is commonly found.

ASSOCIATION OF IRIS PIGMENTATION AND
BIOGEOGRAPHICAL ANCESTRY
[0277] This Example demonstrates that cryptic population structure as determined using AIMS allows an inference as to a complex genetic trait such as iris color.
[0278] In order to determine whether and how common polymorphisms are associated with natural distributions of iris colors, 851 individuals of mainly European descent were surveyed at 335 SNP loci in 13 pigmentation genes and 419 other SNPs distributed throughout the genome and known or though to be informative for certain elements of population structure. Numerous SNPs, haplotypes and diplotypes (diploid pairs of haplotypes) were identified within the OCAS, MYOSA. TYRPI, AIM, DCT and TYl2 genes and the CYPIA2-15q22-ter, CYPIBI-2p21, CYP2C~-1Oq23, CYP2C9-1Oq24 and MAOA-Xpl 1.4 regions as significantly associated with iris colors. Half of the associated SNPs were located on chromosome 15, which corresponds with results others have previously obtained from linkage analysis. Five additional genes (ASIP, MCIR, POMC, and SILV) and one additional region (GSTT2-22q11.23) with haplotype and/or diplotypes were identified, but not individual SNP alleles associated with iris colors (see, also, Intl. Publ.
No.
WO 02/097047). For most of the genes, multilocus gene-wise genotype sequences were more strongly associated with iris colors than haplotypes or SNP alleles.
Diplotypes for these genes explain 15% of iris color variation. These results provide a comprehensive candidate gene study for variable iris pigmentation, and constitute a classification model useful for the inference of iris color from DNA. The results further demonstrate that cryptic population structure can serve as a leverage tool for complex trait gene mapping if genomes are screened with the appropriate AIMs.
[0279] Iris pigmentation is a complex genetic trait that has long interested geneticists, anthropologists, and public at Large, but is not yet completely understood.
Eumelanin (brown pigment) is a light-absorbing polymer synthesized in specialized melanocyte lysosomes called melanosomes. Within the melanosomes, the tyrosinase (TYR) gene product catalyzes the rate-limiting hydroxylation of tyrosine to 3,4-dihydroxyphenylauine, or DOPA, and the resulting product is oxidized to DOPA quinone to form the precursor for eumelanin synthesis.
Though TYR is centrally important for this process, pigmentation in animals is not simply a Mendelian function of TYR or any other single protein product or gene sequence. In fact, study of the transmission genetics for pigmentation traits in man and various model systems suggests that variable pigmentation is a function of multiple, heritable factors whose interactions appear to be quite complex (see, e.g., Akey et al., supj°a, 2002; Box et al., J. Invest. De~matol. 116, 224-229, 2001). For example, unlike human hair color (Storm et al., Gefze 277:49-62, 2001), there appears to be only a minor dominance component for mammalian iris color determination (Brauer and Chopra, Ahthnopol Anz. 36:109-120, 1978), and there exist minimal correlation between skin, hair and iris color within or between individuals of a given population. In contrast, between-population comparisons show good concordance; populations with darker average iris color also tend to exhibit darker average skin tones and hair colors. These observations suggest that the genetic determinants fox pigmentation in the various tissues are distinct, and that these determinants have been subject to a common set of systematic and evolutionary forces that have shaped their distribution in the world populations.
[0280] At the cellular level, variable iris color in healthy humans is the result of the differential deposition of melanin pigment granules within a fixed number of stromal melanocytes in the iris. The density of granules appears to reach genetically determined levels by early childhood and usually remains constant throughout later life, though a small minority of individuals exhibit changes in color during later stages of life.
Pedigree studies suggested iris color variation is a function of two loci; a single locus responsible for de-pigmentation of the iris, not affecting shin or hair, and another pleiotropic gene for reduction of pigment in all tissues (Brues, Amen. J. Phys. Anthropol. 43:387-91, 1975).
[0281] Most of what is now known about pigmentation has been derived from molecular genetics studies of rare pigmentation defects in man and model systems such as mouse and D~osophzla. For example, dissection of the oculocutaneous albinism (OCA) trait in humans has shown that many pigmentation defects are due to lesions in the TYR gene, resulting in their designation as tyrosinase (TYR) negative OCAs (see, e.g., Getting and Ding, Hum.
Mutat. 13: 99-115, 1999; see, also, Albinism database, on world wide web ("www"), at URL
"cbc.umn.edu/tad/"). TYR catalyzes the rate-limiting step of melanin biosynthesis and the degree to which human irides are pigmented correlates well with the amplitude of TYR
message levels. Nonetheless, the complexity of OCA phenotypes has illustrated that TYR is not the only gene involved in iris pigmentation. Though most TYR- negative OCA
patients are completely de-pigmented, dark-ixis albino mice (C44H), and their hwnan type IB
oculocutaneous counterparts exhibit a lack of pigment in all tissues except for the iris (Schmidt and Beermann, P~°oe. Natl. Acad. Sci. USA 24;91:4756-4560, 1994). Study of a number of other TYR-positive OCA phenotypes showed that, in addition to TYR, the oculocutaneous 2 (OCA2; Durham-Pierre et al., Nature Gehet. 7:176-179, 1994;
Durhazn-Pierre et al., Hum. Mutat. 7:370-373, 1996; Gardner et al., Hum. Mutat. 7:370-373, 1992), tyrosinase lilce protein (TYRPl; Boissy et al., Amen. J. Hum. Genet. 58:1145-1156, 1996), melanocortin receptor (MCIR; Robbins et al., supy~a, 1993; Smith et al., supra, 1998;
Flanagan et al., supra, 2000) and adaptin 3B (AP3B; Ooi et al., EMBO J.

16:4508-4518, 1997) loci, as well as other genes (reviewed by Sturm et aL, supra, 2001) are necessary for normal human iris pigmentation. Each of these genes is part of the main (TYR) human pigmentation pathway.
[0282] In Drosophila, iris pigmentation defects have been ascribed to mutations in over 85 loci contributing to a variety of cellular processes in melanocytes (Ooi et al., supra, 1997), but mouse studies have suggested that about 14 genes preferentially affect pigmentation in vertebrates (reviewed in Strum et al., supra, 2001), and that disparate regions of the TYR and other OCA genes are functionally distinct for determining the pigmentation in different tissues. Human pigmentation genes break out into several biochemical pathways, including those for tyrosinase enzyme complex formation on the inner surface of the melanosome, hormonal and environmental regulation, melanoblast migration and differentiation, the intracellular routing of new proteins into the melanosome and the proper transportation of the melanosomes from the body of the cell into the dendritic arms towards the keratinocytes.
Nonetheless, the study of human OCA mutants suggests that the number of phenotypically active pigmentation loci is manageably small for genetic analysis.
[0283] Though research on pigment mutants has made clear that a small subset of genes is largely responsible for catastrophic pigmentation defects in mice and man, it remains unclear whether or how common SNPs in these genes contribute towards (or are linked to) natural variation in human iris color. A brown-iris locus was localized to an interval containing the OCA2 and MYOSA genes (Eiberg and Mohr, Eu~. J. Hum. Genet. 4:237-241, 1996), and specific polymorphisms in the MCI R gene are associated with red hair and blue iris color in relatively isolated populations (see, e.g., Robbins et al., sups°a, 1993; Flanagan et al., supra, 2000; Valverde et al., Natur°e Genet. 11:328-330, 1995; Schioth et al., Biochem. Biophys. Res.
Cornnz. 260:488-491, 1999). An ASIP polymorphism was reported to be associated with both brown iris and hair color (I~anetsky et al., Aruer~. J. Hug. Gehet. 70:770-775, 2002).
However, the penetrance of each of these alleles appears to be low and, in general, appears to explain but a very small amount of the overall variation in iris colors within the human population (Spritz et al., Natuf°e Gev~et. 11:225-226, 1995). However single gene studies have not provided a sound basis for understanding the complex genetics of human iris color.

[0284] Because most htunan traits have complex genetic origins, wherein the whole oiften times is greater than the sum of the parts, innovative genomics-based study designs and analytical methods for screening genetic data i~ silico are needed that are respectful of genetic complexity - for example, the multi-factorial and/or phase known components of dominance and epistatic genetic variance. The first step however is to define the complement of loci that on a sequence level explain variance in trait value, and of these, those that do so in a marginal, or penetrant sense will be the easiest to find. It is towards this goal that the present study was performed.
[0285] A non-systematic, hypothesis-driven genome screening approach was applied to identify various SNPs, haplotypes and diplotypes marginally (i.e., independently) associated with iris color variation. As disclosed in this Example, a surprisingly large number of polymorphisms in a large number of genes were associated with iris colors, indicating that the genetics of iris color pigmentation axe quite complex. The sequences that were identified provide the basis of a classifier model for the inference of iris colors fiom DNA, and the nature of some of these as markers of BioGeographical Ancestry has implications for the design of other complex trait gene mapping studies.
METHODS
Specimen Collection (0286] Specimens for re-sequencing were obtained from the Coriell Institute in Camden, New Jersey. Specimens for genotyping were of self reported European descent, of different age, sex, hair, iris and slcin shades and they were collected using informed consent guidelines under IRB guidance. Donors checked a box for blue, green, hazel, brown, black or unlcnown/not clear iris colors, and each had the opportunity to identify whether iris color had changed over the course of their lives or whether the color of each iris was different.
Individuals for whom iris color was ambiguous or had changed over the course of life were eliminated from the analysis.
[0287] For 103 of the subjects, iris colors were reported using a number from 1-11 as well, where 1 is the daxkest brown/black and 11 is the lightest blue identified using a color placard.
For these subjects, digital photographs of the right iris were obtained, where subjects peered into a box at one end, at the camera at the other to standardize lighting conditions and distance, and from which a judge assigned the sample to a color group.
Comparing the two, 86 of the classifications matched. Of the 17 that did not, 6 were brown/hazel, 7 were green/hazel and 4 were blue/green discrepancies though none were gross discrepancies such as brown/green, brown/blue or hazel/blue. Though such an error is tolerable for identifying sequences marginally associated with iris colors, confidence can be increased for use of the sequences described herein for iris color classification by obtaining digitally quantified iris colors.
SNP Discovery [0288] Candidate SNPs were obtained from the NCBI:dbSNP database, which generally provided more candidate SNPs that were possible to genotype. Human.
pigmentation and xenobiotic metabolism genes were examined, selected based on their gene identities not their chromosomal position. For some genes, the number of SNPs in the database was low and/or some of the SNPs were strongly associated with iris colors, warranting a deeper investigation.
For these genes, xe-sequencing was performed; of the genes disclosed herein, 113 SNPs were discovered in the CYPIA2 (7 gene regions, 5 amplicons, 10 SNPs found), CYP~C~
(9 gene regions, 8 amplicons, 15 SNPs found), CYP2C9 (9 gene regions, 8 amplicons, 24 SNPs found), OCA2 (16 gene regions, 15 amplicons, 40 SNPs found), TYR (5 gene regions, amplicons, 10 SNPs found) and TYRPI (7 gene regions, 6 amplicons, 14 SNPs found; see Tables 9 and 10; see, also, Intl. Publ. No. WO 02/097047).

Candidate Genes Tested for Sequence Associations with Human Iris Pigmentation Gene Name Homology/Model Phenotype AP3B1 adaptor-related proteinmouse "pearl"

complex 3, beta 1 subunithuman HPS2 ASIP agouti signaling proteinmouse "agouti"

DCT dopachrome tautomeraseTYR-related protein mouse "slaty"

MC1R melanocortin 1 receptormouse "extension"
(e) OCA2 oculocutaneous albinismmouse pink-eyed dilution II (p) SILV silver homologue mouse "silver"
(si) TYR tyrosinase mouse "albino"
(c), Himalayan TYRP1 tyrosinase related mouse "brown" (b) protein 1 MYOSA myosin VA mouse "dilute"
(d) POMC proopiomelanocortin mouse Pomcl AIM membrane associated mouse "underwhite"
(uw) (MATP or AIM-1)transpot-ter protein AP3D1 adaptor-related proteinmouse "mocha" (mh) complex 3, delta 1 subunit RAB RAB27A oncogene mouse "ashen" (ash) [0289] Resequencing for these genes was performed by amplifying the proximal promoter (avg. 700 by upstream of transcription start site), each exon (avg. 1400 bp), the 5' and 3' ends of each intron (including the intron-exon junctions, average size about 100 bp) and 3' UTR
(avg. 700 bp) sequences from a multiethnic panel of 672 individuals (450 individuals from the Coriell Institutes DNA Polymorphism Discovery Resource, 96 additional European Americans, 96 African Americans and 10 Pacific Islander, 10 Japanese and I O
Chinese; this 672 represented a separate set of samples than that used for the association study described herein). PCR amplification was accomplished using pfu TURBO polymerase according to the manufacture's guidelines (Stratagene). A program was used to design re-sequencing primers in a manner respectful of homologous sequences in the genome, to insure that pseudogenes were not co-amplif ed or that sequences from within repeats were amplified.
BLAST searches confirmed the specificity of all primers used. Amplification products were subcloned into the pTOPO" sequencing vector (Invitrogen) and 96 insert positive colonies were grown for plasmid DNA isolation (the use of 670 individuals for amplification step reduced the likelihood of an individual contributing more than once to this subset of 96 selected).
[0290] Sequencing was perform with an ABI3700 sequencer using PE Applied Biosystems BDT chemistry; sequences were deposited into a commercial relational database system (iFINCH, Geospiza; Seattle WA). PHRED qualified sequences were imported into the CLUSTAL X alignment program and the output of this was used with a second program to identify quality-validated discrepancies between sequences. Those sequences for which at least two instances of PHRED score 24 or greater variants were identified were selected, and each of these SNPs discovered through re-sequencing was used for genotyping.
Genotyping:
[0291] For most of the SNPs, a first round of PCR was performed on the samples using the high-fidelity DNA polymerase pfu TURBO polymerase and the appropriate resequencing primers. Representatives of the resulting PCR products were checked on an agarose gel, and first round PCR product was diluted" then used as template for a second round of PCR. The two rounds were necessary because many of the genes queried were members of gene families; the SNPs resided in regions of sequence homology and the genotyping platform required short (approximately 100bp) amplicons. For the remainung samples, only a single round of PCR was performed. Genotyping was performed for individual DNA
specimens using a single base primer extension protocol and an SNPstreamTM 25K/Ultra High Throughput (UHT) instrument (Orchid Biosystems; Princeton NJ). Genotypes were subject to several quality controls; two scientists independently pass/fail inspected the calls, requiring an overall UHT signal intensity greater than 1,000 for >95% of genotypes and clear signal differential between the averages fox each genotype classes (i.e. clear genotype clustering in 2-D space using the UHT analysis software).
Statistical Methods [0292] To test the departures from independence in allelic state within and between loci, the MLD exact test was used (Zaylcin et al., supf°a, 1995). Haplotypes were inferred using the haplotype reconstruction method (Stephens et al., supra, 2001). To determine the extent to which extant iris color variation could be explained by various models, Ra values were calculated for SNPs, Haplotypes, and Multilocus Genotypes data by first assigning the phenotypic value for blue eye color as 1, green eye color as 2, hazel eye color as 3, and brown eye color as 4. BGA admixture proportions were determined as described (Hams et al., supoa, 1986; Shriver et al., supra, 2003) within the context of a software program developed for this purpose. For R2 computation, the following function was used:
Adj-R~= 1- f (n/(n-p)}(1-R2), where n is the model degrees of freedom and n-p is the error degrees of freedom. To correct for multiple tests, the empirical Bayes adjustments for multiple results method was used (Steenland et al., Cancer Epiden2iol. 9:895-903, 2000, which is incorporated herein by reference).
RESULTS
[0293] To identify SNP loci associated with variable human pigmentation, 754 SNPs were genotyped, including 335 SNPs within pigmentation genes (AP3B1, ASIP, DCT, MCIR, OCA2, SILV, TYR, TYRPl, MYOSA, POMC, AIM, AP3Dl and RAB, see Table 9), and 419 other SNPs distributed throughout the genome. Alleles fox these latter SNPs were informative for certain elements of population structure; 71 were selected from a screen of the human genome based on their exceptionally high 8 values (i.e., exceptional AIMS) for IndoEuropean, sub-Saharan African, Native American and East Asian BGA (see Example 2;
SEQ ID NOS:1 to 71; see, also, Shriver et al., supra, 2003), and the rest were found in or around xenobiotic genes, which tend to exhibit dramatic sequence variation as a function of BGA. Genotypes for these 754 candidate SNPs were scored for 851 European derived individuals of self repoz-ted iris colors (292 blue, 100 green, 186 hazel and 273 brown).
[0294] Before screening these genotypes for association with iris colors, the 71 non-xenobiotic metabolism ATMs were used to determine BGA aehnixture proportions for each sample, and were tested for correlation between BGA admixture and iris colors. This test showed that each of the 851 Caucasian samples was of majority IndoEuropean BGA, and, though 58% of the samples were of significant (>4%) non-IndoEuxopean BGA
admixture, there was no correlation between low levels (less than 33%) of East Asian, sub-Saharan African or Native American admixture and iris colors, and no correlation between higher levels (greater than 33% but lower than 50%) of Native American admixture and iris colors; there was, however, a weak association between higher levels of East Asian and sub-Saharal African admixture and dancer iris colors.
[0295] It was unclear from the outset whether better success would be realized by considering iris color in terms of 4 colors (blue, green, hazel and brown) or groups of colors.
One method of grouping colors is light = blue + green and dark = hazel +
broom, and this grouping seems to more clearly distinguish individuals with respect to the detectable level of eumelanin (brown pigment). Given that iris color data was self reported, partitioning the sample into brown and not-brown, or blue and not blue could provide greater power to detect significant associations, particularly for alleles associated with only one color. To take advantage of each of these 4 methods, all were considered when screening SNPs for associations; the ~ value, chi square and exact test p-values were calculated for a) a114 colors, b) shades, using light (blue and green) vs. dark (hazel and brown), c) blue vs. brown, and d) brown vs. not-brown (blue, green and hazel) groupings. Significance levels were fixed at 5%, and the alleles of 20 SNPs were associated with specific iris colors, 20 with iris color shades, 19 with blue/brown color comparisons and 18 using the brown/not brown comparison. The overlap among these SNP sets was high but not perfect; a SNP
with a significant p-value for association using at least one of the four criteria is indicated as "marginally" associated.
[0296] When multiple simultaneous hypotheses are tested at set p-values there is the possibility of enhanced type I error. As such, a correction procedure was used to compensate for this risk (Steenland et al., supy~a, 2002); most of the associations were significant after this correction. Most of the marginally associated SNPs were within the pigmentation genes -OCA ~ (11 SNPs at the level of colors), TYRPI (3 SNPs at the level of colors), MYOSA (2 SNPs at the level of colors), AIM (3 SNPs at the level of colors) and DCT (2 SNPs at the level of colors) - though some associations were found within non-pigmentation genes such as CYP2Cg at 1Oq23, CYP2C9 at 1Oq24, CYP1B1 at 2p21 and MAOA at Xp11.3 - also referred to as marginally associated SNPs. No significant SNP associations were found within the pigmentation genes SILY, MCIR, ASIP, POMC, RAB or TYR, though TYR
had one SNP with a p=0.06. The most strongly associated of the marginally associated SNPs were from the OCA2, TYRPI and AIM genes, in order of the strength of association.
[0297] Since most of the SNPs identified from this approach localized to discrete genes or chromosomal regions, all of the SNPs from each locus were grouped and inferred haplotypes were tested for association with iris colors using contingency analysis. This higher-order analysis was not confined to those genes with marginal SNP associations, but grouped SNPs for all of the genes tested. For each gene, haplotypes were inferred and contingency analyses was used to determine which haplotypes were statistically associated with iris colors. From the chi-square and adjusted residuals, 43 haplotypes for 16 different loci were either positively (agonist) or negatively (antagonist) associated with iris colors (Table 10). The strongest associations were observed for genes with SNPs that were marginally associated ;
most of these genes had haplotypes and diplotypes (sometimes referred to as multilocus gene-wise genotypes or diploid pairs of haplotypes) positively (agonist) or negatively (antagonist) associated with at least one iris color (Table 10). A few of the genes/regions not harboring a marginally associated SNP had haplotypes and diplotypes either positively and/or negatively associated with iris colors (ASIP gene - 1 haplotype, MCI R gene-2 haplotypes, Table 10). In other words, their SNPs were only associated with iris colors within the context of gene haplotypes or diplotypes. For some, associations with iris colors were only found within the context of diplotypes, but not at the level of the SNPs or haplotype (i.e. SlL
Tl and GSTT2-22q11.23).
Table 10 The common haplotypes and diplotypes for the 16 iris color genes Gene and Sequence Agonist2 Agoiust2 Antagonist2 Antagonist2 ID. Haplotypes-16 Genesl Chi P Color Chi P Color Count3 AIM
ASIP
DCT

1 C A C 0.0010 Blue 0.004 Bxov~m (1641) 2 T G T 0.048 Brown 0.003 Blue (33) 3 T A C ___ ___ ___ ___ (23) 1 A T A ___ ___ ___ ___ (979) 2 A T G ___ ___ ___ ___ (508) 3 G C G ___ ___ ___ ___ (196) 4 A C A 0.017 Hazel --- --- ( 13 ) 1 C T G A C A --- --- --- --- (625) 2 C T C A C A 0.014 Hazel 0.028 Blue (242) 0.016 Blue 0.021 Browl2 3 T C G A C A 0.048 Green 0.003 Hazel (281) 4 C T C G T A --- --- --- --- (320) C T G A C G --- --- --- --- ( 179) 1 T C C 0.016 Green --- --- (152) Gene and Sequence Agonist2Agonist2Antagoiust2Antagonist2 ID. Haplotypes-16 GeneslChi Color Chi P Color Count3 P

2 C C C --- --- 0.026 Green (1294) 3 C C T ___ ___ ___ ___ (143) 4 C T C ___ ___ ___ ___ (113) MYOSA .

1 C G A T C G G C C C 0.000 Green --- --- (SI) 2 CGATCAGCCC --- --- _-- ___ (gsg) 4 GTGCTGATCC --- --- 0.008 Blue (163) C G A T C A A C C C 0.017 Blue --- --- (40) 6 G T A C T G A T C C --- --- --- --- (117) 8 C G A C T G G T T T --- --- --- --- (165) C G A C C A G C C 0.003 Brown --- --- (40) C

13 C T A C T G G T T --- --- _-- --- (71 T ) 14 C G A T T A G C C --- --- 0.027 Blue (40) C

16 C G G C C A A T C --- --- --- --- ( 19) C

GGGGACGGCA 0.002 Blue 0.01 Brown 1 A A G (44) GAGGCCGGCA 0.018 Hazel 0.031 Blue 2 A G A (43) G A G G C C A G C A 0.022 Brown 0.026 Blue 3 A G A 0.042 Green (21) GAGGCCGGCA 0.024 Brown 0.014 Blue 4 A A A (g9) GAGGCCAGCA --- --- -_- ---7 A A A (13) T G G G A C G C T A --- --- --- ---AAG (17) GAGGCCAGCG 0.019 Brown --- ---19 A G A (23) GAGGCCGGCG 0.036 Green 0.012 Blue 22 A A A (15) T A A G C C A G C G --- --- --- ---A A A (13) G G A A A T A G C A <0.001 Blue <0.001 Brown 37 A A A (508) GGAAATAGCG 0.025 Blue 0.011 Brown 38 A A A (200) GGAAATAGCA --- --- 0.039 Brown 39 A G A (71) GGAAATAGCA --- --- _-- ---4I GAA (19) Gene and Sequence Agonist2AgonistaAntagonistzAntagonistz ID. Haplotypes-16 Chi Color Chi P Color Count3 Genes P

G G A A A T A 0.029 Blue 0.021 Brown G C G

42 A G A ( 174) T G A A A T A <0.001 Brown 0.007 Blue G C G

45 A A A (65) T G A A A T A 0.003 Brown 0.006 Hazel G C G

47 A G A (22) T G A A A T A 0.001 Brown 0.036 Blue G C A

48 A A A 0.015 Green (35) TGAAATAGCA --- --- ___ ___ 57 A G A (30) POMC

1 T ___ ___ ___ ___ (1187) 2 C ___ ___ (515) 2/2 C/C 0.043 Blue --- ___ SILV

1 T C __, ___ ___ ___ (913) 2 C C __, ___ ___ ___ (296) 3 T T --- (490) 1 / 1 TC/TC --- --- 0.03 5 Blue TYR

1 G C G --- --- 0.01 Green (184) 2 A A G ___ ___ ___ ___ (257) 3 G C A ___ ___ ___ ___ (467) 4 A A A ___ ___ ___ ___ (231) A C A ___ ___ ___ ___ (254) 6 A C G ___ ___ ___ ___ (189) 7 G A G ___ ___ ___ ___ (83) 8 G A A --- --- 0.023 Blue (37) TYRP

1 T T T T C G 0.007 Blue <0.001 Brown (968) 2 C T T T T T ___ ___ ___ ___ (86) 3 C T G A C G 0.006 Brown --- --- (454) 4 C C G A C G 0.046 Brown 0.029 Blue (9g) 5 T T G T C G --- --- --- --- (23) 7 C T T A C G ___ ,__ _-_ ___ (24) CYP 1 , 1 Sq22-ter ~

1 G 0.015 Brown 0.023 Hazel (769) 2 C 0.023 Hazel 0.015 Brown (933) Gene and Sequence Agonist2Agonust2Antagonist'Antagonist2 ID. Haplotypes-16 Chi P Color Chi P Color Count3 Genesl 2p21 1 C C ___ ___ ___ ___ (957) 2 T T 0.027 Hazel 0.025 Blue (403) 3 C T ___ ___ ___ ___ (337) 1Oq23 1 C A A 0.007 Brown --- --- (539) 2 T A G ___ ___ ___ -__ (20I
) 3 T G A --- --- 0.039 Brown (513) 4 T A A --- --- --- --- (439) 1Oq24 1 T 0.023 Brown 0.039 Green (1325) 2 C 0.039 Green 0.023 Brown (377) 22q11.23 1 A G ___ ___ ___ ___ (821) 2 G A ___ ___ ___ ___ (778) 3 A A ___ ___ ___ ___ (g9) AG/AG 0.040 Green --- --- (184) AG/GA 0.013 Hazel --- --- (401) MAOA-Xp11.4-11 1 C G --- --- 0.027 Blue (1133) 2 T A 0.006 Blue 0.026 Hazel (480) 3 C A ___ ___ __- ___ (87) 1 Sequences of the highest order of complexity within a locus found to be associated with iris colors. All of the major sequences (count > 13 ) for each locus with at least one significantly associated sequence are shown. If no haplotypes or diplotypes for a locus were found to be associated, only the SNP alleles are shown. If no haplotypes were found to be associated for a locus but diplotypes were found to be associated, both the haplotypes and diplotypes are shown.
2 Agonist color refers to the color with which the sequence is positively associated.
Antagonist color refers to the color with which the sequence is negatively associated.
Chi-square P-value is shown.
3 Number of times the haplotype was observed in our sample of 851.

[0298] At the level of the haplotype, each gene or region had unique numbers and types of associations. For example, OCA2, AIM, DCT and TYRPI harbored haplotypes both positively associated with blue irides and negatively associated with brown hides (OCA2 haplotypes 1, 37, 38, 42, AIMhaplotype 1, DCT haplotype 2, and TYRPI haplotype 1 Table 10). Others genes such as AIM, OCA2 and TYRPI harbored haplotypes that were positively associated with brown but negatively associated with blue color (AIMhaplotype 2, OCA2 haplotypes 2, 4, 45, 47, TYRP haplotype 4, Table 10), while others such as MYOSA, OCA2, TYRPI and CYP2C8-1Oq23 harbored haplotypes that were positively associated with one color but not negatively associated with any other color (MYOSA haplotype 5, haplotype 10, OCA2 haplotype 19, TYRPI haplotype 3 and CYP2C8-1Oq23 haplotype 1, Table 10).
The MCI R gene harbored haplotypes only associated with green color in our sample and the POMC gene harbored a single SNP with genotypes weakly associated with iris colors (no significant haplotypes or diplotypes were found).
[0299] Overall, the diversity of haplotypes associated with brown irides was similar to that of haplotypes associated with blue hides. Most of the haplotypes were even more dramatically associated with iris colors in a multi-racial sample, because many of the SNPs comprising them are good AIMS, and variants associated with darker iris colors were enriched in those ancestral groups of the world that are of darker average iris color. Most of the SNPs within a gene or region were in LD with others in that gene or region (D'<0.1 ); only 32 SNP pairs, in the MCI R ( 1 pair), OCA2 (27 pairs), TYR (2 pairs) and TYRPI
(2 pairs) genes were found to be in LD.
[0300] These analyses resulted in the identification of 61 SNPs in 16 genes/
chromosomal regions associated with iris colors on one level or another, whether the SNP
is marginally associated or associated within the context of the haplotype and/or diplotype.
The minor allele frequency for most of these SNPs was relatively high (avg. f minor allele = 0.22) and most of them were in Hardy Weinberg Equilibrium (those for which HWE p>0.05, 28/34, Table 10). Nine were not and of these, 2 were of relatively low frequency and the evidence for disequilibrium was marginal (p value close to 0.05). Lack of HWE is usually an indication of a poorly designed genotyping assay, and none of the remaining 7 SNPs exhibited genotyping patterns that we have previously associated with such problems (such as an absence of one genotype class, or a preponderance of heterozygotes). Indeed one of those for which the evidence of lack of HWE was the strongest was validated as a legitimate SNP
through direct DNA sequencing. The chromosomal distribution of the SNPs that were significantly associated in a marginal sense was independent of the distribution of SNPs actually surveyed, indicating that the associations were not merely a function of SNP
sampling..
[0301] Chromosome 15q harbored the majority (18/34) of the SNPs that were marginally associated with iris colors, and 14 of these chromosome 15 SNPs were found in two different genes OCA2 and MYOSA. Chromosome Sp had 3 SNPs marginally associated, all in the AIM
gene amd chromosome 9p had5 SNPs associated, all in the TYRPI gene. Multiple SNPs were identified on chromosome lOq; the CYP2C8-1Op23.33 region had 2 SNPs, and the neighboring region CYP2C9-1Op24 also had one. All 3 markers were in tight LD
with one another (p<0.001 for each possible pair). Multiple SNPs were also identified on chromosome 2; the POMC SNP located at 2p23 was marginally associated, and SNPs from the CYP 1 B 1-2p2I region were associated within the context of a 2-SNP
haplotype (Table 10), and these SNPs were also in LD (p<0.01 ). Finally, in addition to the OCA2 (15q11.2-q12) and MYOSA (15q21) sequences, a single SNP (15q22-ter),was also implicated on chromosome 15q, but SNPs between each of these three loci were not in LD.
SNPs for the MCIR (16q24), SILh(12qI3), TYR (I lq), MAOA-XpI I.4-I1.3 and GSTT2-22q1I.23 regions were also associated at the level of the haplotype (Tables 10 and 11), though these were the only regions of these chromosomes for which associations were found.
[0302] The p-values that were obtained indicated that diplotypes explained more iris color variation than haplotypes or individual SNPs. To test this, a corrected ANOVA
analysis was performed for the data on each of these three levels. All 61 SNPs were considered, as were their haplotypes (Table 10) and diplotypes (not shown). Diplotypes explained 15°Jo of the variation, whereas haplotypes explained 13% and SNPs explained 11% (Table 4) after correcting for the number of variables. The most strongly associated 68 genotypes of the S43 genotypes observed for the 16 geneslregions, based on chi-square adjusted residuals, explained 13% of the variation (row 4, Table 11).
Table 11 ANOVA- SNP and Haplotype data No.

Row Source Model Error VariablesF ValueR2 DF DF Value Adj.R2 1 SNP data 62 788 62 2.63 0.17 0.11 2 Haplotype data 212 638 216 1.58 0.34 0.13 Multilocus Gene-wise 0.15 3 Genotype data S43 307 572 1.27 0.69 Multilocus Gene-wise 0.13 4 Genotype data 68 782 68 2.82 0.2 [0303] From a screen of 7S4 SNP loci, 61 were identified that were statistically associated with variable iris pigmentation at one level of intragenic complexity or another. The remaining SNPs had 8 values and chi-square p-values that were not significant on any level of intragenic complexity. Diplotypes for these 61 alleles explained most of the iris color variance in the sample; the lowest amount was explained at the level of the SNP, suggesting an element of intragenic complexity to iris color determination (i.e., dominance).
[0304] Only about half of the 61 SNPs identified were associated with iris colors independently; the others were associated only in the context of haplotypes or diplotypes.
Even at this level of complexity, the sequences from no single gene could be used to make reliable iris color inferences, indicating an element of intergenic complexity (i.e., epistasis) for iris color determination as well. Aside from the fact that many of the identified SNPs were significant after imposing the correction protocol for multiple testing, five lines of evidence indicated that the identified SNPs are not spuriously associated.
First, for all of the genes identified as marginally associated SNPs, multiple such SNPs were identified; i.e., the distribution of SNPs among the various genes tested was not random. Second, some of the non-pigment gene SNPs are located near pigment genes, e.g., CYPZC~ (1Oq24.1) and CYP2G9 (1 Oq24), which are located proximal to two pigment genes not tested directly -HPSI (1Oq23.1-4) and HPS6 (1Oq24.34) and the chromosome 2p SNP at the CYP1B1 locus (CYP1B1-2p21) located pxoximal to POMC at 2p23 (and in LD with the POMC SNP).
Third, though a roughly equal number of pigmentation and non-pigmentation gene SNPs were tested, of the 34 marginally associated SNPs, 28 of them (82%) were in pigmentation genes. Thus, the distribution of SNPs among the various gene types also was not random.
Fourth, the associations were generally stronger for the SNPs in the context of within-gene haplotypes, a result that would not necessarily obtain for SNPs spuriously associated (i.e., the result suggests the gene sequences themselves are associated, not merely a single polymoxphiszn within each gene). Fifth, when applied to a sample including individuals of multiple ancestries, the linear and non-linear variables from these and the other genes combined performed even better than when applied just to individuals of majority European ancestry. Since most individuals of non-European, or minority European descent exhibit low variability in iris colors (on average of darker shade than individuals of European descent) this improvement may not seem surprising. However, this result would not have necessarily been obtained were the SNPs not truly associated with iris colors.
[0305] Though corrections for multiple testing left most of the SNP-level associations intact, a number of the associations did not pass the multiple-testing examination, but are presented in order to avoid possible type II error; the sequences may be weakly associated with iris colors and possibly relevant within a multiple-gene model for classification (i.e.; epistasis). For these, it would seem more prudent to eliminate false-positives downstream of SNP identification, such as from tests of higher order association, using various other criteria such as those described above, or possibly using the utility of the SNP
for the generalization of a complex classif canon model.
[0306] Mutations in the pigmentation genes are the primary cause of oculocutaneous albinism, so it was natural to expect that common variations in their sequence explain some of the variance in natural iris colors and, in fact, this result was observed.
However, a number of the associations were for SNPs located in other types of genes (CYP2C8 at l Oq23, CYP2C9 at 1Oq24, CYP1B1 at 2p21 and MAOA at Xp11.3). The inclusion of non-pigmentation genes in this study was intentional; the screen was not restricted to pigmentation gene SNPs, but included two types of AIMs - those that were selected from the genome based on 8 values between sub-Saharan, IndoEuropean, Native American and East Asian population allele frequencies, and those selected based on their location within xenobiotic metabolism genes. The latter are included based, in part, on previous evidence indicating that xenobiotic metabolism genes harbor an unusual concentration of AIMs, and that some of these AIMs are relevant for the measurement of "cryptic"
population structure, pxesumably because xenobiotic metabolism gene products are responsible for detoxification of floral alkaloids and tannins present in indigenous diets, and selection and genetic drift have shaped the geographical distribution of their sequences. Such cryptic structure may be correlated with iris colors to a degree that enables accurate classification, even though they may not be helpful for elucidating a biological mechanism.
[0307] It was hypothesized that a) some of these SNPs would be indicators of not merely crude or continental population structure, but sub-structure and perhaps even microstructure, b) iris colors were correlated with these elements of structure within the Caucasian group, and c) these markers can serve as proxies for phenotypically active loci for the purpose of classification or trait value inference. The commonly held notion is that genetic screening is only propexly conducted towards identifying phenotypically active loci through linkage disequilibrium. However, when classification is the goal, rather than the identification of phenotypically active loci, population structure can be helpful if the trait value correlates with the structure and if markers for the structure can be identified. For example, an iris color classification tool can be useful to a forensic scientist for the objective and science-based construction of a partial physical profile from crime scene DNA. Currently, forensics investigators construct physical profiles using surprisingly unscientific means; only in xare cases are eye-witness accounts available, and often times human accounts are subj ective and unreliable in certain circumstances. Fox a forensics application, an investigator is less interested in the biological mechanism of the phenotype than in an ability to make an accurate inference of trait value. Of course, identifying markers in LD with phenotypically active loci (or the phenotypically active loci themselves) provide for more accurate classification, as well as for a better understanding of biological mechanism, but the hunt for these elusive loci in heterogeneous populations is impractical because LD only extends fox a few I~b and expensive genome-wide scans are required.

[0308] That a number of the SNPs identified herein as associated with iris colors were located in xenobiotic metabolism genes suggests the identified markers are associated with ixis colors through correlation with cryptic population structure. In other words, the non-pigmentation gene markers are probably correlated with, but not necessarily in LD with, phenotypically active loci for iris color. Through such a correlation, both markers and active loci are enriched for in certain branches of the IndoEuropean pedigree, even though they may not be in LD with one another. These results based on such correlations are meaningful in a classification context only with respect to the sample used. For example, AIMS
selected based on their inter-continental b values were not associated with iris colors in individuals of mainly European descent, but were strongly associated with iris colors in a more international sample because the AIMS, are specifically relevant for the element of structure correlated with iris colors in this sample. In contrast, these same AIMs were not associated with iris colors within the sample of majority European-derived individuals examined because there is little variation in crude structure witlun this sample. Instead, within any (majority) Caucasians or European American sample, there will exist sub-structure or micro-structure (cryptic structure) due to variation in ethnic or other sub-population level affiliations, and only those SNPs specifically xelevant for measuring this cryptic structure would be needed if the structure correlates with the trait. It bears noting that no systematic structure in the mainly Euxopean-derived study sample unrelated to the phenotype measured was identified, indicating that the use of qualified AIMS is imperative for reproducing the present results with another trait.
[0309] Other interpretations of the present results are possible, for example, that the associations could have been observed through LD with as of yet to be defined pigmentation genes. Indeed, CYP2C~ and CYP2C9 axe located on chromosome 10 near the HPSI
and HPS2 pigmentation genes (not tested directly), CYPIA2 is located at 15q22-ter, on the same arm as OCA2 and MYOSA, CYPI BI is located at 2p21, in the vicinity of the POMC
gene at 2p23 and MAOA is located on the same arm of the X chromosome (Xpl 1.4-11.3) as the OAl pigmentation gene (not tested directly). The distances between these loci associated with iris colors and "neighboring" pigmentation genes is far greater than the average extent of LD in the genome, and even if these associations are through LD, it would seem that, again, population structure would need to be invoked as an explanation.
[0310] LD is known to extend fox megabases~in recently admixed populations, such that as few as a couple thousand AIMS can be used to gain full genome coverage in these populations, and it is of some interest that two-thirds of the European American samples used in this study were of significant (4%) BGA admixture. Though European Americans are not recognized as a traditionally defined admixed group (like Hispanics or African Americans), the BGA admixture observed may be linked to finer, cryptic levels of population structure.
While the relevance of the present results to LD and/or population structure is not clear, if the results are due to LD rather than correlation, they would suggest that just as AIMS can be used to leverage population admixture for trait mapping in recently and extensively admixed populations, they also can be used to leverage cryptic population structure in a similar manner. Thus, regardless of whether the results are due to correlation or LD, the large number of non-candidate gene associations identified indicates that the measurement of population structure has broader implications for the cost-effective development of pharmacogenomics and complex disease gene classifiers.
[0311] Linkage studies have implicated certain pigmentation genes as specifically relevant for pigmentation phenotypes, and most of the pigmentation gene SNPs identified herein clustered to certain genes such as OCA2, MYOSA, TYRPI and AIM. Further, certain aspects of the present support the previous literature. Most of the SNPs identified were on chromosome 15, which linkage analyses identified as the primary chromosome for the determination of "brownness" (Eiberg and Mohr, Eu~. J. Hum. Genet. 4:237-41, 1996); it was suggested that the candidate gene within the interval containing this locus (BEY~) was most likely the OCA2 gene, though the MYOSA gene is also present within this interval and, as disclosed herein, was associated with iris colors. OCA2 associations were by far the most significant of any gene or region tested, while MYOSA SNPs were only wealdy associated (but haplotypes and diplotypes more strongly). MYDSA alleles were not in LD
with those of OCA2, suggesting these results were independently obtained and that results described by Eiberg and Mohr may have been a reflection of the activity of two separate genes (Eiberg and Mohr, sups°a, 1996).
[0312] Two OCA2 coding changes were reported to be associated with darker iris colors (Rebbeclc et al., Cahcef Epidemiol. Bioma~ke~s P~ev..11(8):782-784, 2003). In addition, the "red hair/blue iris" SNP alleles previously (Valverde et al., supra, 1995;
I~oppula et al., supra, 1997) were identified, confirming that these sequences are associated with iris pigmentation, though the previously described associations were with blue irides and at the level of the SNP, whereas, in the present study, the associations were with green irides and only apparent at the level of the haplotypes and diplotypes. Associations were also identified in the ASIP gene (see Kanetsky et al., supra, 2002), though, in the present study, this gene association was not at the level of the SNP; one of the ASIP SNPs identified herein (marker 86I) is the 8818 G-A SNP transversion described as associated with brown iris colors (Kanetslcy et al., supra, 2002) though, in the present study, the association was with hazel color at the level of the haplotype.
[0313] The associations between TYR haplotypes and iris colors was relatively weak, which is not inconsistent with results obtained by others in the field of oculocutaneous albinism who have failed to find strong associations in smaller samples.
Though the present results independently verified findings for OCA2, ASIP and MCI R, they also show that several other pigmentation genes harbor alleles associated with the natural distribution of iris colors (TYRPI, AIM, MYOSA and DCT). As such, the present results indicate that most of the previous studies associating pigmentation gene alleles with iris colors, taken independently, represent mere stTOlces of a larger, more complex portrait.
[0314] Interestingly, most of the SNPs identified herein are non-coding, either silent polymorphisms or residing in the gene proximal promoter, intron or 3'UTR. This result, while not altogether unusual, can indicate that the SNPs are in LD with other phenotypically active loci, or it may be a reflection that variability in message transcription and/or turnover may explain part of the variability observed in human iris colors. Though a large number of SNPs was screened, some of the genes harbor a large number of candidate SNPs and not all were tested. For example, the OCA2 has about 200 known candidate SNPs in NCBI
dbSNP.

As such, the OCAZ gene may yet have more information variable human iris pigmentation, such information being accessible using methods as disclosed herein.

USE OF AIMS TO PREDICT DRUG RESPONSIVENESS
[0315] This Example demonstrates that AIMS can be used to develop chemopredictive and diagnostic tests because many drug response traits, as well as many human genetic traits, correlate with elements of population structure.
[0316] The distribution of SNPs available for genotyping by chromosomal arm is shown in Figure 12. At each of the approximately 400 SNPs examined, Caucasian individuals taking Lipitor~ (180), for whom response was known in terms of cholesterol (TC), low density lipoprotein (LDL), liver trarisaminase ASTSGOT and ALTGPT
measurements, were genotyped. 150 ZocorTM patients of known response in terms of TC and LDL
change, and 1,000 individuals of lcnown hair and eye color, also were genotyped. Those SNPs with delta values of significance (b>0.20) among the various trait classes were selected.
For example, in about 70% of patients, LipitorTM caused a decrease in LDL, whereas in 30%
of patients it had no effect. For any given SNP, the delta value (8) is the difference in minor allele frequency among those individuals for whom LDL decreased by at least 20%
versus those for whom LDL did not so decrease. The ~ value was measured for each SNP in Figure 12 for each test (ZocorTM: LDL, TC, ASTSGOT, ALTGPT response; LipitorTM: LDL, TC
response). Fox eye color, the 8value was measured in terms of dark (hazel or brown) versus light (blue or green) eyes. For hair color, the cS value was measured in terms of black or brown color versus blonde.
[0317] For Lipitor~ response, the number of SNPs of significant (8>0.20) value for each of the 4 endpoint measures is shown in Figure 13. Those SNPs with delta values of significance for ZocorTM response using LDL and TC endpoint measures then were selected (Figure 14). Next, those SNPs with delta values of significance for Iris color were selected (Figure 15); and, likewise, for hair color (Figure 16). The distribution of SNPs with good 8 values was similar for each trait in Graphs A-E, but with certain elements of specificity.

The specificity can be appreciated by focusing on chromosome arm 6p, which has many important SNPs for TC (Total Cholesterol) response to LipitorTM (Figure 13), but none for ZocorTM response(Figure 14). Chromosome 2 harbors SNPs with good 8 values for eye color (Figure I 6), but not hair color (Figure 15). Chromosome I S harbors many markers predictive of LipitorTM responsiveness, but not ZocorTM response. This specificity is likely a function of linlcage disequilibrium with other loci deterministic for each of these traits, but which has nothing to do with the expression of the other traits; this type of finding is the goal of traditional measures of gene mapping. Alternatively it can be due to correlation with certain elements of population structure [0318] The similarity is apparent in that chromosomes 10 and 22 have a relatively high number of SNPs with good b values for each of the four mechanistically unrelated traits, as does chromosome 1. Overall, the distribution of important SNPs for one trait is not dissimilar to that for another. It is the similarity between the profiles that is of interest here, and that illustrates the value of the present methods.
[0319] The elements that the four graphs share in common correlate with the number of SNPs genotyped (Figure 12; Figures 13 to 16 roughly resemble Figure 12). At first pass, this result appears to indicate that the :importantness" or significance for these SNP alleles are spurious, and merely a function of the number of SNPs genotyped for each chromosomal arm (i.e., the more SNPs one genotypes from a chromosome, the more SNPs of good delta value one will find on that chromosome). However, the SNPs in Figure 12 are not just any type of SNP; most of the SNPs available in the Figure 12 are xenobiotic metabolism and pigmentation gene SNPs, and most all are good AIMS.
[0320] For iris color, it was shown that most of the SNP associations remain significant (chi p<0.05 after correction for multiple testing, thus indicating that the SNP associations axe not spurious. That the distribution of SNPs associated for each of the 4 traits resemble one another to a large extent, and that this distribution is similar to the distribution of available SNPs, most of which are good AIMs, indicate that most of the SNPs measured in these experiments are reporters of population structure, and that similar elements of population structure are correlated with response to each of the two drugs (however measured) as well as to hair and iris color pigmentation.
[0321] It is notable that these four traits appear to be mechanistically unrelated (at least to current lmowledge) and it is not intuitively obvious as to how hair or eye color can relate to the response to two randomly selected drugs. However, the similarity in profiles for important SNPs for each of the traits suggests that each can be predicted, to a signficant extent, with a knowledge of the sequence for a common set of chromosomal markers.
Because these markers are known to be excellent indicators of BGA, the results indicate that each of the four unrelated traits can be predicted to some extent by measuring BGA rather than by measuring the particular SNPs measured in Figures 12 to 16.
[0322] Simply measuring BGA as disclosed herein does not impart as much predictive power for each of the four traits as measuring the specific AIMS in the plots above.
However, using the markers in Figure 15 enabled a good classification accuracy for iris color.
These results indicate that different AIMs in the plots above are informative for different elements of population structure or substructure, and while most of the SNPs in Figure 12 are good indicators of crude population structure in terms of continental BGA, it has not yet been determined to what extent each is also informative for other finer levels of structure, such as between Scandinavian and Mediterranean ancestry, or even within ethnic groups.
Such "cryptic" structure could not previously be defined in a reliable, confidence qualified manner because, prior to the present disclosure, it was not possible to define the subtle elements of structure recognized in large populations of individuals of common race in biogeographically meaningful ways). For example, red haired individuals of maj ority IndoEuropean descent are known to require 20% more anesthesia than other IndoEuropean (or other) patients (Cohen, supra, 2002). These red haired individuals also exhibit a tendency towards hypertension and bleeding under the influence of certain common anesthetics, thus presenting a serious clinical problem of unknown etiology.
[0323] Red haired individuals are common in the Uuted Kingdom (Ireland and Britain), which also contains individuals sharing more ancestors with one another than with individuals of other regions of Europe. As such, they can be considered as a branch of indeterminate structure off the family tree of humanity. Only one gene is linked to red hair color (MCIR), and it is difficult to imagine this gene is of such pleiotropic effect that its sequence contributes to variability in such diverse and complex physiological responses as those to numerous anesthetics. Further, not every person with red hair harbors the known red hair MC1R variants, yet most exhibit the aberrant anesthesia response. Rather, it is more likely that red hair color correlates with an element of population structure that also correlates with anesthesia response; i.e., the genes) for red hair color and aberrant anesthesia response are unique to or enriched for in particular branches of a human pedigree, within which these two traits are more common. Thus, the genes for red hair color and aberrant anesthesia response are distributed as a function of population structure, and, similarly, so have many other traits, as disclosed herein.
[0324] The AIMs and methods disclosed herein are suitable for the measurement of various levels of structure, including crude continental sti-uCture, as well as structure relevant for ethnicity and even cryptic structure (e.g., almost 30 markers have been identified that are capable of resolving Pacific Islanders from other East Asians; i.e., a finer level of structure than continental structure). The informativeness of the AIMS arises over the course of evolutionary human development through founder effects, migration, bottlenecks, genetic drift and/or selection, but these forces need not be focused on hair color or anesthesia response for there to have arisen AIMs correlated with these two types of traits. Much like this case, where there must be AIMS that are informative for disparate phenotypes in North Western Continental Europeans, there also must be AIMs that are informative for disparate phenotypes in IndoEuropeans in general. The results demonstrating the similarity in AIMS
with good delta values for LipitorTM response, ZocoxTM response, hair color and eye color (Figures 12 to 16) are indicative of a level of population structure that is informative for each of these phenotypes.
[0325] It should be noted that the magnitude of association does not match between the traits. For example, the strongest AIM for iris color is not the strongest AIM
for LipitorTM
response, etc. Also, the direction of the association is not necessarily the same among the traits; thus, agonist associations (associated positively) for blue iris color can be either agonist or antagonist associations (associated negatively) for a particular LipitorTM
response outcome.
[0326] Randomly selected AIMS that distinguish specifically between Africans and East Asians can, but need not, harbor information about a particular set of traits because they can, but need not, be marlcers of the particular element of human population structure correlated with the trait (i.e., the branch of the human family tree within which these traits are more common). Similarly, AIMS that distinguish between Indo Europeans and Africans would not necessarily carry with them the information necessary to help with the prediction of anesthesia response or red hair. Only those AIMS with alleles that distinguish between individuals of the particular branches of humanity within which the traits values are concentrated or underrepresented carry the information necessary to predict trait value, in this example, anesthesia and red hair, or, as disclosed, response to LipitorT~, ZocorTM, hair color and eye color. Certain SNPs axe good AIMs for certain gross elements of population structure (Europeans versus sub-Saharan African) or sub-substructure (Northern Europeans versus Mediterranean IndoEuropeans) or micro-structure (Scottish vs. Irish vs.
British; or red haired Northern Europeans versus Northern Europeans of other hair colors; or Northern Europeans that respond to a drug versus other Northern Europeans that do not respond to the drug). Some AIMS are informative for some levels of population structure but not others, and most of the SNPs in the human genome do not carry information of any level of population structure at all (i.e. they are not AIMs).
[0327] The primary element of the disclosed methods is that most human traits can be predicted through a detailed measurement of AIMS associated with population structure at various levels, provided the trait correlates with that element of structure.
A secondary element is that classifiers, or collections of SNP markers, and methods for predicting trait value from DNA can be constructed for most human traits through such an appreciation of population structure. As disclosed herein, such applications can be accomplished through correlation, not just through extended LD found in certain admixed groups such as Hispanics or African Americans, but for any sample of subjects, regardless of race or ethnic baclcground, provided that the AIMS used are appropriate for the element of population structure with which the trait is correlated. These results can be attained because the methods of the invention provide a means to mine the genome for good AIMs, qualify their value as AIMS and accurately measure population structure against the backdrop of human phenotypes.
[0328] The trend observed in the studies represented by Figures 12 to 16 has been observed for many other traits, where the SNP associations axe of such "penetrance" that they withstand corrections for multiple tests (Steenland et al., supra, 2000). As such, by measuring AIMs across the human genome, elements of population structure, sub-structure, or microstructure relevant for predicting or inferring the value of virtually any given human trait can be learned, even if the markers are not in linkage disequilibrium with the phenotypically active loci for the trait. This correlation applies to any human trait, simple or complex, of clinical, recreational, forensics or other value, or not, because every human trait is correlated to greater or lesser degree with certain elements of population structure.
[0329] It is not possible to know a p~°iof°i whether a trait segregates with population structure versus substructure versus sub-substructure versus microstructure unless one measures AIMS across the genome in individuals and identifies which (if any) correlate with trait value. Virtually any trait correlates with at least one element of structure, some with gross structure (as in the case of human shin pigmentation), some with substructure (as in the case of iris, hair or slcin pigmentation between South Asian IndoEuropeans (Indians) and Northern European IndoEuropeans (e.g., Irish), and some with microstructure (as in the case of red hair or anesthesia response among Continental Europeans). It is not important to know with which level of population structure a trait correlates in order to measure and find AIMs for the inference of that trait, it is only necessary to measure and test a plurality of general AIMs for statistical association with that trait. As such, methods are provided for linking human gene sequences with traits such that they can be predicted or inferred.
Such a method is valuable, for example, in the clinical and forensic fields because it is not important that biologically or mechanistically relevant sequences be measured when predicting a common trait (response to a drug or predisposition to develop disease), as it is to make an accurate inference of that common trait (drug response or disease disposition).

ALGORITHM FOR MAXIMUM LIKELTHOOD ESTIMATE
[0330] An software program was written based on the algorithm of Harris et al.
(supra, 1986) for using multilocus AIM genotypes to determine the Maximum Likelihood Estimate of individual BGA admixture. A flow chart illustrating an algorithm useful for determining proportional ancestry is provided in Table 12. An example as to how the algorithm operates in shown in Table 13, and the results of an ancestry proportion calculation is shown in Table 14.
[0331] The 8 value is an expression of the ancestry informativeness of the maxker (Dean et al., 1994). For a biallelic marlcer, the frequency differential (~) is equal to pX - py which is equal to qy- qX, where pX and py are the frequencies of one allele in populations X and Y and q~ and qy are the frequencies of the other. To test the departures from independence in allelic state within and between loci, we have used the MLD exact test (Zaykin et al., supra, 1995).
The collection of 71 AIMs used in Example 2 was selected to maximize the cumulative 8 value within, and minimize differences in the cumulative 8 value between each of the six possible pairs of the four dimensional (sub-Saharan African, Native American, IndoEuropean and East Asian) problem.
[0332] The algoritlun inverts the population specific allele frequencies to obtain a likelihood estimate of proportional affiliation corresponding to a multilocus genotype using three groups at a time; three groups were used mainly for computational convenience, and also because a 4-dimensional admixture is lilcely to be relatively rare. For example, the likelihood of 100% TndoEuropean, 0% Native American, 0% East Asian is calculated, then the likelihood of 99% IndoEuropean, 1% Native American, 0% East Asian is calculated next, and so on until all possible IndoEuropean, Native American and East Asian proportions axe considered, and then the process is repeated for all possible IndoEuropean, Native American and Afiican proportions, and all possible Native American, African and East Asian proportions. The likelihood of maximum value is selected as the Maximmn Likelihood Estimate (MLE). When plotting a single MLE on a triangle plot, the space within which the likelihood is within 2-fold, 5-fold and 10-fold of the MLE also are delimited (see Figure 3);
these intervals generally are not plotted when multiple MLEs are shown in a single triangle plot.

Algorithm for Ancestry Calculation - Flow Chart Calculate Allele Frequencies for all SNPS of given populations For each population Likelihood Value for the population (LVP) = 0 T-~~ For All Markers I I Expected gen type value (EGV) = 0 I T ~--- YES ---~ if SNP has heterozygote NO ~
I
I I EGV = log (2*frequencies of allelel EGV = log (frequencies of allelel I I in the population * in specified population I I frequency of allele2 frequency of allele2 I I in the population ) in the population ) I T I I
I- ~ -.
i I I Likelihood Value for Population (t_VP)=LVP+EGV
T I
I I I
I
I I
Select 3 populations, which have Maximum Likelihood Values Maximum Lilcelihood Value (MLV)=0; Ancestry Proportion for populationl (API) =0;
Ancestry Proportion for population2 (~AP2) =0; Proportion for population3 (AP3) =0;
For 7~1 = 0 to 100 T I ' ~ For ~,2 = 0 to 100-~,1 I I
I T ~,3=100-~,1-~,2 Likelihood Value for possible combination (LV)= 0 i For All Markers Expected genotype value (EGV) = 0 Allelel Estimated Freguency (A1 EF) _ 7~1 * Allelel frequency in population .T + ~,2 * Allelel frequency in population 2 + ~,3 * Allelel frequency in population 3 YES ~ if SNP has heterozygote NO ~
I I I Allele2 Estimated Frequency (AZEF) = I
I I I ~1 * Allele2 frequency in population 1 + I
I I I 7~~ * Allele2 frequency in population 2 + I
I I I ~.3 * Allele2 frequency in population 3 .L
T ~r T .~ I
EGV = log (2*A1EF*A2EF) EGV = log (A1EF*A1EF) I ~ I
I I I I I
i Lv=LV+EGv I:
I E NO ~-if ( (7~3 = 100 ) or (LV > MLV) ) I YES
I
I MLV=LV; API=7~1;AP2=~.2; AP3=~,3 I
E
~,1, ~,2, 7~3 are corresponding Ancestry proportions for Blind Sample.

Example of Proportional Ancestry Determination Using Algorithm SAMPLEID . SNP1 SNP 2 I. Pick best three population for blind sample II. Get proportions which has maximum likelihood value i. Pick up best three populations:
Algorithm:
For all populations f For all snps f populationSum <- populationSum + Expected genotype frequency.
Picl~ three populations, which have maximum value.
STEP 1:
For all population STEP2:
SNP1 has heterogeneous genotype; Alleles are G and T.
Expected genotype for SNP1 = log ( 2*P(G,1)*P(T,1));
SNP2 has Homogenous genotypes Expected genotype for SNP2 =log (P(T,1)*p(T,1));
Likelihood value population) = Expected GT for SNPi+Expected TT for SNP2 Repeat STEP 2 for all population Pick best three likelihood values out of those 4 population values.
For those selected three population estimate proportions.

1. Starts with ~1=0, ~2=0 and ~3=1 0+0+1= 1 2. Compute Likelihood value:
Estimate Expected genotype for SNPI
SNP1 has heterogeneous genotype; Alleles are G and T.
Estimated Allele Frequency from SAMPLE:
Allelel Estimated Frequency (AIEF)= ~,l.p(G,1)+7~z. p(G,2)+~,3. p(G,3) Where p(G,1) - G Allele frequency in population i p(G,2) - G Allele frequency in population 2 p(G,3) - G Allele frequency in population 3 The mixing proportions ~,1, ~z ana ~3 are treated as unknown parameters.
Allele2 Estimated Frequency (A2EF)=~,~. p(T,1)+~,z, p(T,2) +~,3. p(T,3) Where p(T,1) - T Allele frequency in population 1 p(T,2) - T Allele frequency in population 2 p(T,3) - T Allele frequency in population 3 The likelihood of the parameters is obtained by multiplying the probabilities for each observed genotype in the new observation under the assumption of Hardy-UVeinberg La w.
Since SNP1 has heterogeneous genotype Expected genotype for SNPI = log(2*A1EF*A2EF);
Estimate Expected genotype for SNP2 Estimated Allele Frequency from SAMPLE:
Allelel Expected Frequency (AIEF)=~,~, p(T,1)+ ~,z. p(T,2) + ~,3. p(T,3) Where p(T,1) - T Allele frequency in population 1 p(T,2) - T Allele frequency in population 2 p(T,3) - T Allele frequency in population 3 The mixing proportions 7~1, ~,z and ~3 are treated as unknown parameters.
Since SNP2 has Homogenous genotypes Expected genotype for SNP2 = log(A1EF*A1EF);

LIKELIHOOD VALUE
Compute LIKELIHOOD VALUE by adding all Expected genotypes of all SLAPS
LIKELIHOOD= Expected genotype for SNP1 + Expected genotypes for SNP2;
Compute LIKELIHOOD value by using different unknown parameters. (Repeat step 2) 3. Get Maximum Likelihood value and corresponding unknown parameters.
Those unknown parameters are nothing but proportions Ancestry Proportion Calculation:
Ancestry Freauencv Tabie African frequencies SNP1 G 0.8 T 0.2 SNP2 T 0.7 A 0.3 SNP3 C 1.0 European frequencies SNP1 G 0.9 T 0.1 SNP2 T 0.7 A 0.3 SNP3 C 0.8 T 0.2 Native (NA) American frequencies SNP1 G 0.6 T 0.4 SNP2 T 0.5 A 0.5 SNP3 C 0.7 T 0.3 Middle East (ME) frequencies SNP1 G 0.7 T 0.3 SNP2 T 0.9 A 0.1 SNP3 C 0.9 T 0.1 EXAMPLE

III. Pick best three population fior blind sample IV. Get proportions which has maximum likelihood value a. Pick three best populations African(Assume blind sample 100% African):

SNP1 Alleles: G, T

"G" allele frequency in AfricanP(G) = 0.8 "T" allele frequency in AfricanP(T) 0.2 =

Expected genotype value for - log (2 * P(G) SNP 1 * P(T)) - log (2*0.8*0.2) - - 0.4948 SNP2 Alleles: T, T

"T" allele frequency in AfricanP(T) 0.7 =

Expected genotype value for - log(P(T) * P(T)) - log (0.7 * 0.7) - - 0.3098 SNP3 Alleles: C, T

"C" allele frequency in AfricanP(C) 0.9999 =

"T" allele frequency in AfricanP(T) 0.0001 =

Expected genotype value for - log (2 * P(C)*
SNP3 P(T)) - Iog(2*0.9999*0.0001) - - 3.6990 Likelihood for African = - .6990 0.4948 - 0.3098 - 3 - - 4.5036 Euronean(Assume blind sample I00% European)~
SNP1 Alleles: G, T
"G" allele frequency in European P(G) = 0.9 "T" allele frequency in European P(T) = 0.1 Expected genotype value for SNPl - log (2 * P(G) * P(T)) - log (2*0.9*0.1) - - 0.7447 SNP2 Alleles: T, T
"T" allele frequency in European P(T) = 0.7 Expected genotype value for SNP2 - log(P(T) * P(T)) - log (0.7 * 0.7) - - 0.3098 SNP3 Alleles: C, T
"C" allele frequency in European P(C) = 0.8 "T" allele frequency in European P(T) = 0.2 Expected genotype value for SNP3 - log (2 * P(C)* P(T)) - log (2*0.8 * 0.2) - - 0.4948 2. Lilcelillood for European - - 0.7447 - 0.3098 - 0.4948 - - 1.5493 Native American(Assume blind sample 100% NA):
SNP 1 Alleles: G, T
"G" allele frequency in NA P(G) = 0.6 "T" allele frequency in NA P(T) = 0.4 Expected genotype value for SNP1 - log (2 * P(G) * P(T)) - log (2 * 0.6 * 0.4) - - 0.3187 SNP2 Alleles: T, T

"T" allele frequency in NA P(T) = 0.5 Expected genotype value for SNP2 - log(P(T) * P(T)) - log (0.5 * 0.5) - - 0.6020 SNP3 Alleles: C, T
"C" allele frequency in NA P(C) = 0.7 "T" allele frequency in NA P(T) = 0.3 Expected genotype value for SNP3 - log (2 * P(C)* P(T)) - log (2*0.7 * 0.3) - - 0.3767 3. Lil~elihood for Native American - -0.3187 - 0.6020 -0.3767 -1.2974 Middle East(Assume blind sample 100% MEI:
SNP1 Alleles: G, T
"G" allele frequency in ME P(G)= 0.7 "T" allele frequency in ME P(T)0.3 =

Expected genotype value for - log (2 * P(G) SNP 1 * P(T)) - log (2*0.7*0.3) - - 0.3767 SNP2 Alleles: T, T

"T" allele frequency in ME P(T)0.9 =

Expected genotype value for - log(P(T) * P(T)) - log (0.9 * 0.9) - - 0.0915 SNP3 Alleles: C, T

"C" allele frequency in ME P(C)0.9 =

"T" allele frequency in ME P(T)0.1 =

Expected genotype value for - log (2 * P(C)*
SNP3 P(T)) - log (2*0.9 *
0.1) - - 0.7447 4. Likelihood for Middle East - -0.3767 - 0.091 S - 0.7447 - -1.2I29 Likelihood value for African = - 4.5036 5. Likelihood value for European = - 1.5493 6. Likelihood value for Native American = - 1.2974 7. Likelihood value for Middle East = -1.2129 In this case we drop African and consider other three for proportions.
Maximum likelihood value Now starts giving some values to unknown parameters I = European J= Native American K = Middle East Always I + j + lc =1 I=0; j=0; lc=1 f SNP1 Alleles: G, T.
"G" allele frequency in European P(G,1 ) = 0.9 "G" allele fiequency in NA P(G,2) = 0.6 "G" allele frequency in ME P(G,3) = 0.7 Allelel Estimated Frequency(AlEF) = I * P(G,1) + J*P(G,2) + K* P(G,3) = 0'* 0.9+0*0.6+ 1*0.7 = 0.7 "T" allele frequency in European P(T,l) = 0.1 "T" allele frequency in NA P(T,2) = 0.4 "T" allele frequency in ME P(T,3) = 0.3 Allele 2 Estimated Frequency(AIEF) = I * P(T,I) + J*P(T,2) + K* P(T,3) = 0*0.1+0*0.4+1*0.3 = 0.3 Expected genotype value for SNPl = log (2*AlEF*A2EF) = log (2*0.7*0.3);
- 0.3767 SNP2 Alleles: T, T.
"T" allele frequency in European P(T,1) = 0.7 "T" allele frequency in NA P(T,2) = 0.5 "T" allele frequency in ME P(T,3) = 0.9 Allele 1 Estimated Frequency(AIEF) = I * P(T,1) + J*P(T,2) + K* P(T,3) = 0*0.7+0*p.5+1*0.9 = 0.9 Expected genotype value for SNP2 (EGV2) = log(AlEF*A2EF) = log (0.9*0.9);
- - 0.0915 SNP3 Alleles: C, T.
"C" allele frequency in European P(C, 1) = 0.8 "C" allele frequency in NA P(C, 2) = 0.7 "C" allele frequency in ME P(C, 3) = 0.9 Allele 1 Estimated Frequency(AlEF) = I * P(C,l) + J*P(C,2) + K* P(C,3) =0*0.8+0*0.7+1*0.9 = 0.9 "T" allele frequency in European P(T, l ) = 0.2 "T" allele frequency in NA P(T,2) = 0.3 "T" allele frequency in ME P(T,3) = 0.1 Allele 2 Estimated Frequency(AlEF) = I * P(T,1) + J*P(T,2) + K* P(T,3) = 0 * 0.2 + 0*0.3 + 1 *0.1 = 0.1 Expected genotype value for SNP3(EGV3) = log (2*AlEF*A2EF) = log (2*0.9*0.1);
- - 0.7447 i. Lil~elihood value for unlcnoWn parameters = EGV 1+EGV2+EGV3 - -0.3767 - 0.0915 - 0.7447 - -1.2129 for European = 0; NA=0; MiddleEast =l; likelihood value is -1.2129 Repeat above loop for all possible combinations 0.0,0.0,1.0 -1.2129 1.0, 0.0, 0.0, 0.0, 1.0, 0.0 , 0.1, 0.0 0.9, 0.1, 0.1,0.8, O.I,0.2, 0.7, 0.1,0.3,0.6, 0.1,0.4, 0.5 etc Get Maximum likelihood value and corresponding proportions axe ancestry proportions.
[0333] Although the invention has been described with reference to the above example, it will be understood that modifications and variations are encompassed within the spirit and scope of the invention. Accordingly, the invention is limited only by the following claims.

SEQUENCE LISTING
<110> DNA Print Genomics, Inc.
FRUDAKIS, Tony N.
SHRIVER, Mark D.
<120> COMPOSITIONS AND METHODS FOR INFERRING ANCESTRY
<130> DNA1170-2W0 <150> US 60j404,357 <151> 2002-08-19 <150> US 60j467,613 <151> 2003-05-02 <160> 363 <170> PatentIn version 3.1 <210>

<211>

<212>
DNA

<213> Sapiens Homo rs236919 <400>

tgagcgtggaggtggagtgggaagatgggtcaacttgcagcagttctccygggactgtat 60 gaagagggcatcccgtttctcagagtctttgtgcaggccagttacctttccaagctgatg 120 ggggccatccactgtcttcccatcgtcatctggtccccagctgttgagaaataaatacct 180 agaggggacagccagagccaggctgcggctacccacccagcgcctcccagtctcagctgg 240 ctgccctttg 250 <210> 2 <211> 170 <212> DNA
<213> Homo Sapiens rs594689 <400> 2 gccgggcaga gacagcggct agagaccccg aggtggggga cacaaatgar ctccttgccc 60 gtccccccaa tttctcctac cctacaaagg gagggaatcc caaaaggacc cttctacccc 120 tgtgactggt tttcatgaag gactatgatt cccatcatcc ctcaggggca 170 <210> 3 <211> 189 <212> DNA
<213> Homo Sapiens rs725192 <400> 3 caaatccctg aaagagtaag cattatgcta atgggtctga aaatataaay gaagaatcaa 60 aatctattct gctttaaaga aggcaccagt ttaggaggca ggctctaagt acttgaagca 120 ataaaaggaa ctcccctttg cctaacaagt cccatatagc agcctccttc atacattcag 180 tttaacgtc 189 <210> 4 G211> 110 G213> Homo Sapiens rs1800498 G400> 4 aggaatgatg cctggatgcc tgggggtgtg aagaaaagag ccttgggtty gactagggaa 60 cctggggcca ctccttcctc tgctactaaa tcaccaagtg atcttgttct 110 G210> 5 <211> 170 G212> DNA
G213> Homo sapiens rs593226 G400> 5 cttccttcta tgacccctac ctgttgtggg tctccaggca agcagtgcay aggtagagcc 60 atccttaggt agcctttaga cttaatatta ggtgagctct ccccacagat agcctctcct 120 ttatttgaat ggaattatat tttaagtttg gaaatatttt tcagcttatt 170 <210> 6 <211> 170 <212> DNA
<213> Homo Sapiens rs1337038 <400> 6 ctctcaaaca ttagctttta cccttttacc ttcaggttcg taccccaccr tatttgcata 60 attgtattac ttccctaaga accaattttt tgttgttgtt gtttgctagt acttttgtgc 120 cacctgacaa acttacaagc tatttttaaa tgcttgcttt taaatagagg 170 <210> 7 <211> 126 <212> DNA
<213> Homo Sapiens rs729531 <400> 7 ccagtaggtg tcagctacgc acagcctttc tctttctaaa tccatatttr taagaaaacg 60 gcctttcatc cataattttc cctagttaaa gggatcaatt tcttctgatc tgaagcacgc 120 aaaaca 126 <210> 8 <211> 170 <212> DNA
<213> Homo sapiens rs1111108 <400> 8 gaaacgccat ggctgttctc ataccaagtg ctatttaata agacagatcr ctcaaactgg 60 gatcccaaga gggttcaatt aagagcataa tgtaattcca ggctaccagt gctgccctca 120 gcccaagtat ctcatacctg etctatgaag tctgtcctga tgtccccaga 170 <210> 9 <211> 170 <212> DNA
<2l3> Homo Sapiens rs1007407 <400> 9 tggcctgtga agtccagaag ctctttccgc gctcggctct ttgagcctcr gtttccccat 60 ctgtcaggcg tgtgtatgtg ggaggggggt acggagtaga tttacctaga tcactcgggg 120 ttgtcccttc cagttttgag ggtgatgtcg gggtgtctgt gaggaggagg 170 <210> 10 <211> 170 <212> DNA
<213> Homo Sapiens rs741272 <400> 10 gtcgctgtca gttgaaatag attacaagag ggatctcctg gcagaatttr tttagtttag 60 ctcaaaggag aatgttatag tcaagggtac atttggttgc aaggacaaga gagtcactcg 120 agttggctta agtaaaaggg tacttgtccc aaaaacataa cagccagtcc 170 <210> ll <211> 170 <2l2> DNA
<213> Homo Sapiens rs1076808 <400> 11 aatctctttt ccagaagtct aagtgaaaat ctcactttgt ctctgtactr gttaggcaag 60 gctaagttgc tatgacaaat agacctcaaa ttctgctgta acagctcaga ctataacact 120 gcattttgct gtcatgtaat atcagtggtg ggcctttctg gtcatttctc 170 <210> 12 <211> l70 <212> DNA
<213> Homo Sapiens rs716873 <400> 12 ctccatgctg catgtaggtt tataaaccaa taccctgctc agacaccacr ttcttctaac 60 accctaatga tggggtcaca tttctctctt taggtactac aatctggttt tgaagttgcc 120 caaagaaact tgaatggctt cctctttcat tcccacaagg ctctgaaaac 170 <210> 13 <211> 170 <212> DNA

<213> Homo Sapiens rs730570 <400> 13 gacactcccc tactccaggc ctgcagcact cacctgcatc tcacactgcr atgcaaaatt 60 tgtgtgatta atgggttcaa atgtgccccc ctcccctgga ctttacaggg tgctggagct 120 gctggcttgc tgttcccagt gccaagtccc atatgggtgc tcagccatat 170 <210> l4 <211> 311 <212> DNA
<2l3> Homo Sapiens 886993 (OCA2E10 102) <400> 14 ttaatatggt gtcctgctaa aagattatcc ttgtcttctt cttttccccr tagatgatct 60 tagtagccat attttcagaa acgggatttt tcgattattg tgctgtaaag gtaggtatga 120 tgttgcattt aataattcta tcctgattaa tttatatatg tatttttctg acattatata 180 tttaggaaac aaacatttaa aacaaacatt gaaaaattcc atccttcttt taaaggttgc 240 ttctgcaggg cagggtatac ttgctatgtt aagttgtatg gctctgagca gcactttcag 300 ctgctcagta a 3l1 <210> 15 <211> 242 <212> DNA
<213> Homo Sapiens 886895 (OCA2Ell 263) <400> 15 cctcttcacg cctgtgacca taaggtacgc aaagcacctc tgccgtgggr gttgcggcca 60 ggttctggca ggcaggggct ctgcctgcac tgcctggctc caggttccat tctcaggtgc 120 atgaaaaggt gggggcrgtt gagcccacag ctcactgcat tccagtccag ctcgtgtctg 180 ctttgtgtga ctgcagtaca tgctacaagc agtggggcct cagaagctgg tggcagaaat 240 gc 242 <210> 16 <211> 210 <212> DNA
<213> Homo Sapiens rs1426208 <400> 16 gcctgtaaag acccttctct tgctttggaa acaccctccc tctcctgccr tgatcaaccc 60 tgatgttggg tagttctgca gtgggcacgt ctgtacatgc ctacccccaa 110 <210> 17 <211> 95 <212> DNA
<213> Homo Sapiens rs2862 <400> 17 tctgagtgca agataaaaag gaaatagcaa ttcaaggcac aaagctaagy acatgcaaca 60 atatagatga tttgggggtg ggacagtaca gaatt 95 <210> 18 <211> 110 <212> DNA
<213> Homo Sapiens rs972801 <400> 18 ttcttctgaa tctagatgca gcatttagca actctatact attcaccaay attatcttgc 60 tttatgtttt caacctataa aattcaactc agtttctgaa aaaggtgtag I10 <210> 19 <2l1> 110 <2l2> DNA
<2l3> Homo Sapiens rs251741 <400> 19 gaatctgctg agactgacgc tggcgtccca accagacact tggaggcatr gggcatgccc 60 ttcaagtcta gaaccctgcc aggtacctat ccctcctgcc caggccaaag 110 <210> 20 <211> 73 <212> DNA
<2l3> Homo Sapiens rs1650999 <400> 20 catatggtgc aataaaattg tggtacagca aagacaaaga gaaaaaccay agaaagaacc 60 gcagtaaaaa cat 73 <210> 21 <211> 110 <212> DNA
<213> Homo Sapiens rs173537 <400> 21 ccctcccctc ttgggtcttg tagtcggctg tgggtggcaa gtacagcagy gatggcagac 60 aggtgtgatg agttcaagga cagggaaata acatccacag ccccatccca 110 <210> 22 <211> 64 <212> DNA
<213> Homo Sapiens rs67302 <400> 22 tgtcacctcg gatcacccgg gaacgtagac tccaaactcc caatggctty agcagaaatt 60 gage 64 <210> 23 <211> 110 <212> DNA
<213> Homo Sapiens rs212498 <400> 23 gggtgaatgg atggatgagt gctgggaggc tgtattgtaa ataacaaaar tgggatctga 60 aggaaggtat gggactctga agtctgcatt ccttttacct tttgctatat 110 <210> 24 <21I> 110 <212> DNA
<213> Homo Sapiens rs1010923 <400> 24 gataaattga attaaacgtg gtattttatt tggatttctt gatttttaay gtaagcatga 60 attatagtaa ctcttccttg ctatctgaaa tactgaaatt cttgttctag 110 <210> 25 <211> 110 <212> DNA
<213> Homo Sapiens rs430667 <400> 25 gcgctgatgt ggttgtggag gtcactgtgg agaccacaga ggggtagttr tcatgtccag 60 gteccaactg ttaagcatga agggaagtat agttttaaag gcttttgtgt 110 <210> 26 <211> 110 <212> DNA
<213> Homo Sapiens rs236336 <400> 26 gaagcgtccc tgtgggcccc cagaaataac tgtcagaggc atgactggcr ggggagatgg 60 gcaagggtcc tgccattatg aagtgggttc caagagtcag cagagttgag 110 <210> 27 <211> 110 <212> DNA
<213> Homo Sapiens rs783064 <400> 27 cattgttctg tcttggagag ggttcagtgc tgggtgggga gacatctgar aacagattat 60 tgcaattaag tgagaaaagc tgaaagttaa attgcctaga gtgtgcattc 110 <210> 28 <211> 110 <212> DNA
<213> Homo Sapiens rs595961 <400> 28 gccaggggct tttgctcccc tacccacctg actctactga ggctcacctr ggcgccccct 60 ctacctatcc ccagaaccgg gccattgcca cacccaatca gggtgtctgg 110 <210> 29 <211> 110 <212> DNA
<213> Homo Sapiens rs434504 <400> 29 tccccactct ctggctgcat agaggtttta atgaaacaag gggataaccr tctccctcct 60 gctggctctg ttgacactgc cagtgggttt tgcaggcaga gcaccctgcc 110 <210> 30 <211> 110 <212> DNA
<213> Homo Sapiens rs2814778 (FX-null) <400> 30 gcccgcagac agaagggetg ggacggctgt cagcgcctgt gcttccaagr taagagccaa 60 ggactaatga gggccatcag gttctgggca gggacagggg gcatcaggaa 110 <210> 31 <211> 110 <212> DNA
<213> Homo Sapiens rs1325609 <400> 31 tatgcaagag tgctgagaga cgagggatga tggctccact gttgtgctgs ccacattgta 60 tttggtgagt ccaggctcct gctcttatca gaacagcaat ggtagctgct lI0 <210> 32 <211> 110 <212> DNA
<213> Homo Sapiens rs6003 <400> 32 ggttacatct ctgatgtaaa gttattgtat aaaattcaag agaacatgcr ttatggttgc 60 gcttcagggt acaaaaccac tggagggaag gatgaagaag tggttcaatg 110 <210> 33 <2ll> 110 <2l2> DNA
<213> Homo sapiens rs2065160 <400> 33 catctgaaaa cacattgtaa agaatggcct ctcgatgagt aaatatgggy caaatgaagt 60 aaacagacta tgcgtaggta tcatcaaata tagagagaat gaacagaccg 110 <210> 34 <2l1> 110 <212> DNA
<213> Homo Sapiens rs1406869 <400> 34 cagggcccca tattgaaaag agccatccat gctcattagc atggctctcy cctccctcct 60 tatgtgtggt aatcatgtga gaaacagtca gctgagcttg tgacatgaaa 110 <210> 35 <211> 116 <212> DNA
<213> Homo Sapiens rs1004571 <400> 35 gctgtcggga cccctactga gagaatagag accttgtgtt ttatctcttc ttgccrgaga 60 gctctgccta tggcctcgca gacagcaatc agtcaaggct cagccaccca cacttc 116 <210> 36 <211> 110 <212> DNA
<213> Homo sapiens rs172982 <400> 36 agattgcaag caaggcattc tcgagtttca attctggaag agaaagcacr ttggagcatt 60 tggtaggacg aactagctcc agggeaaaga tactgtggcc ggcactcact 110 <210> 37 <211> 110 <212> DNA
<213> Homo Sapiens rs138335 <400> 37 gatgtttaat tcaacatgca gttcagaaaa atgacagatt tgtcttgtas aaaaagacct 60 aacacaagct aagcctttaa gaaaaccaac ctcaactgca tgaggcactg 110 <210> 38 <211> 110 <212> DNA
<213> Homo Sapiens rs3287 <400> 38 attataatcc atcctccaac acacacacaa ataagcagct aatggcaatr ctagtggtct 60 tcccaattca caagacctgt gcttcaaatt gttttcctga taatgtggag 110 <210> 39 <211> 110 <212> DNA
<213> Homo Sapiens rs959929 <400> 39 gtagtgtagt ctttattctt ggtgaaagga aagaaatgtg caaatgctts tttgcctgtt 60 gcacagtttc aaggatgaat ttatccacac atccactttt gaaggtgcct I10 <210> 40 <211> 110 <212> DNA
<213> Homo Sapiens rs716373 <400> 40 ccaggtccct gtggctggac ggggcacagc ttgggagcag ggcttcatgy tacctttgct 60 ctgtgcagca aagccctgca tggtcactgg tgtctggctc tctcctcctc 110 <210> 41 <211> 110 <212> DNA
<2I3> Homo sapiens rs729253 <400> 41 ttgcctatcc tagtaccacc ctcactgctt ctggtgctca cccatttctr gaacaaagct 60 ctgcattttt tatatcaaat gtgcagacgt gtgccacaag cacacctcct 110 <210> 42 <211> 110 <212> DNA
<213> Homo Sapiens rs1108718 <400> 42 actagctccc agaaagtagt ttatcacggc tcaacaaaag atcaataacr ttgattgaat 60 tccattcata ctagaaaaca tatctattta actgtaccct cagaaaaatt 110 <210> 43 <221> IIO
<212> DNA
<213> Homo Sapiens rs1399272 <400> 43 aactttaaac ctattgtcta ccacatccaa ccaccacatt gctttggctr cagagatcag 60 tttaaagagc agatgattaa aagaattaac cacactttag aggtgaagaa 110 <210> 44 <211> 110 <212> DNA
<2I3> Homo Sapiens rs361065 <400> 44 cagtggtgac ttgttagtta tgtgactaca aacaaggcac agaaaactgr ttgagcttca 60 ggtttttcat atccttgaga attatagcat ctgtatgtca gggttattaa 110 <210> 45 <211> llo <212> DNA
<213> Homo Sapiens rs361055 <400> 45 aatagaaaaa ccaccagaat tagctacacg gaatacacgg atcactgccr tctcctcaga 60 gggatcaatg cgtttaagaa aaacacttgg ctcagcccaa ggagaaccag 110 <210> 46 <211> 110 <212> DNA
<213> Homo Sapiens rs733563 <400> 46 ccagcgtaac caccttctcc tgactttagc cctgacatcc ctgctttctr gtaatcccca 60 gctccaggca agcccgaagg ccagcaccct acactcaccc ttcctgggtg 110 <210> 47 <211> 785 <212> DNA
<213> Homo Sapiens 712047 (HMGCRE12E14-3_786) <400>

acccaaaaggtatcactaattgtccttaaaactcttctCattgccttacytatgatgtat 60 ttttaaactggcaaatatataaatgccaacttacacctattgctctgcagcctctattgg 120 tgctggccacaagacaaccttctgttgttgccattggaacctgaaattctttttcatcta 180 agcaaaggggtcctgccactccaacagggatgggcatatatscaataacattctcacaac 240 aagctcccatcacctaaaaggtaaagtcaggcaccaaatgaaaatctatatagtaaatgc 300 acaaaattttatctcagcttgtcagtataactatcttcaaacttaatcctttagtatgta 360 ttctttttaaacaaaatgtttaattcacctttaaaaagtgtttaaaatactaaaatcttt 420 gagtaatgactataaaagcaggaaatatatttttaatttctatgacctacatcataagca 480 gaataaaaaattaggataaaatgatttaaaaggaaaatgttttataaaactatgtacata 540 tatgaaactaatcaagaaaaaagattaaaaacatacaaagcaaatctctcttaatcgaga 600 aaataacatmccaaggagtaattataatccctgtaaggtaggtactggagagaagaaggt 660 tctgaaagcttcttggaaagtaactgtcggcgaatagatacaccacgctcatgagtttcc 720 atcagagtttccaacttgtaggctgggatatgcttagcattgactaactggatgatctca 780 gcatc 785 <210> 48 <211> 110 <212> DNA
<213> Homo Sapiens rs960709 <400> 48 ggccctaagc cagagctgga gctcaggccc ccagagttcc gtccagggcr ctccaggact 60 cttcttaaca aagctataga aattccgagc tgaaaagacc cataaagact 110 <210> 49 <211> 110 <212> DNA
<213> Homo sapi.ens rs3317 <400> 49 atttacatct atttatttcc attcagtgta tcacatcttc aggataggtr ataacagtgt 60 gaagggtgtg ctcattttct tcagctgtga gtagaggagt cttcccgaga 110 <210> 50 <211> 104 <212> DNA
<213> Homo Sapiens rs3340 <400> 50 gaaatcgaat acgtccattt ctttgtaaaa taacaataao gttraaggca aaagcaagat 60 tctgtaaacc aacattggaa aaggggacac agggaggggc agag 104 <210> 51 <211> 98 <212> DNA
<213> Homo Sapiens rs933199 <400> 5l agggaaaggg aagtaggcat gtaccatttt aatgcacrcc ttggttttca gaaaatttaa 60 gaagaaagac tttctgcttt tctctgacta ttctgtat 98 <210> 52 <211> 110 <212> DNA
<213> Homo Sapiens rs1476597 <400> 52 ctgtccctat cttcaccagc ctgccctcaa atcattggtt caaatcgcts tcgagaagtt 60 tgggagctcc cagggcgggg gctttctaac agctctgcct agctcgggcc 110 <210> 53 <211> 110 <212> DNA
<213> Homo sapiens rs1155513 <400> 53 ggaaaagagt gtggaagtta actaaatctg cagctaatac cacagaaccr tggagatgat 60 ctgaatgaag aaacttagaa aatgtggtgg gtcagaacaa gagtttactc 110 <210> 54 <211> 110 <212> DNA
<213> Homo Sapiens rs984654 <400> 54 aaaaaatcct gataagcact gttaggcata ttaactttaa tgattacaay ttttaggaca 60 ctctgtggcc tagacttaga aacacaacta atgtccagaa aaagattcct 110 <210> 55 <211> 110 <212> DNA
<213> Homo Sapiens rs2161 <400> 55 tcccctccct cggcacccac tcccaggcca agagcgtcct atgaatcatr catttgttcc 60 tgttattgct gttcacagag tggcaactct tgcaaaggga ggggtacaaa 110 <210> 56 <211> 51 <212> DNA
<213> Homo Sapiens CYP3A4E10-5 292 <400> 56 ttttacccaa taaggtgagt ggatgrtaca tggagaagga gggaggaggt g 51 <210> 57 <211> 110 <212> DNA
<213> Homo Sapiens rs869337 <400> 57 atccctgcaa cttggcacgt cttgctgtaa atggaagcag tggcatgtcy gtgtcaggca 60 tgaaagccgt tcataaaacc accttgctga gcacttcctg tctgtcaggc 110 <210> 58 <2I1> 110 <212> DNA
<2l3> Homo Sapiens rs285 <400> 58 ctgcaagggt tttgcttaat tctcaattca atgtctcttc atcttttagy agctgtgggg 60 ttttgttgtt gttcttctgt ttttgcttag tatctgacta ctttttaatt 110 <210> 59 <21l> 110 <212> DNA
<213> Homo Sapiens rs716840 <400> 59 acagaagtat tttaactcct attttcattt tattccettt ctaaaactar ctcagatcat 60 agtcaggcga ttgcttcatt ctcagctttg tggagagtga tatacacaga 110 <210> 60 <211> 79 <212> DNA
<213> Homo Sapiens rs3176921(CRH2942) <400> 60 ggagtagetc ttgtcatcat ctaaaaaact tgaactgcat tttgagagay ttattggcct 60 tgcttctgca ggctcataa 79 <210> 61 <211> 110 <212> DNA
<213> Homo Sapiens rs1467044 <400> 61 agaatccccc ttgccattcc catccctgta tggtccttcc tctagggctr tttgggaata 60 gtacttatgg tctaactggt tctgcagctg cccttcgtgt ggttataaat 110 <210> 62 <211> 110 <212> DNA
<213> Homo Sapiens rs998599 <400> 62 ctttcactgt gttttgtcgt gtatgtagga tttacccagt agcaaaactr taaattcagg 60 gtcttoaagc taaaacttcc taaagcaata ccctaagact cagttgtgaa 110 <210> 63 <21l> 110 <212> DNA
<213> Homo Sapiens rs104I321 <400> 63 attctcagtc tgagcatctg aaggaggatg gggactttat cccttcatcy caggtctgta 60 aatgtcgtgt gtatgaatga atccacacca gtccettccc atagacaaca 110 <210> 64 <211> 110 <212> DNA
<213> Homo Sapiens rs1888952 <400> 64 agaagttcac tctccttccc agatcataga cagaagacct gggttcaagy cacgactctg 60 ctacagacct gcctgtgatt tcaggctaac cgctgagggt cacagagata 110 <210> 65 <211> l10 <212> DNA

<213> Homo Sapiens rs913258 <400> 65 atcagaggga gagaataaat gaccctgata ctcttactta ggaaggcats ctcccccctc 60 cctcccctcc ctcgagaatc atggctttaa agacacagag atgagaaaga 110 <210> 66 <211> 82 <212> DNA
<213> Homo Sapiens rs2695 <400> 66 agcatactga agactattcc taaatggcta tttgtgttgg gtggtcaagr ctattcagaa 60 aatctcagag gaggacaaat ga 82 <210> 67 <211> 110 <212> DNA
<213> Homo Sapiens rs675837 <400> 67 ttctttttat ttatgtctaa ctcctcagca aaggaaggcc ttacttaacr tgataaacat 60 tcacatttta ttttagttaa gcatttcttt aaatgtaaga aattggaggc 110 <210> 68 <211> 110 <212> DNA
<213> Homo Sapiens rs662117 <400> 68 ctaaagggga cactactgtg tccaggagca tCtCCtCtCt cccaacacay acacacacaa 60 caaaataaga tcaaacctgc tgatttaata tcaatttttc aacaaaaatt 110 <210> 69 <211> 110 <212> DNA
<213> Homo Sapiens rs590086 <400> 69 catgtttttt atgagtcctc tggcgccttt ttgccttggc taccgaagcr tcttgaaagc 60 agagattttt gaaagcccca ttgatttttg tgtcatccat ctcagatcct 110 <210> 70 <211> 110 <212> DNA
<213> Homo sapiens rs9032 <400> 70 aaaaccagtg cctcttgtaa gcagccgcct caacactgaa cgcagaaaas aggacaaaac 60 catctgcgtg tcagcctgtc agcatatcac tatataaatt aaaagtcaaa 110 <2l0> 71 <211> 110 <212> DNA
<213> Homo Sapiens rs1076160 <400> 71 gccagacctt agtggggaaa gcgtttcaga cccagacagt atgtctcccr catggcatag 60 agaatggctt gatgacttat ttcctaaaag acggagcggc agtcccacta 110 <210> 72 <211> 100 <212> DNA
<213> Homo Sapiens 32934 <400> 72 gcgcacgcac cagagccagt ggcaaagcaa cttaagctgc aggctcctcy tgcccacccc 60 cagccgcgac gcccctgggg gcccttagca gactgccaag 100 <210> 73 <211> 100 <2I2> DNA
<213> Homo Sapiens 235936 <400> 73 atagtatttc ttttcttcag tagaatggtc aggtctctaa gtgctgcagy ggcttgaact 60 cttgatgaca aacatgtagg aaaacttcat gcagccagtt 100 <210> 74 <211> 100 <212> DNA
<213> Homo Sapiens 520354 <400> 74 cttcaaaggc cagaggcttg gagctggcac aaggagtcct tgtgcagaar agaacagaag 60 gtggttttgc tgccctacat ccatccccgg aaccttctca 100 <210> 7S
<211> 100 <212> DNA
<213> Homo Sapiens 539619 <400> 75 tccctataag gtatactgtt taaagactag gtccttggct ttgccttcar gtaacttgaa 60 gatgtttttg ttgaaatatc ttggagacta aaaaatagtc 100 <2l0> 76 <211> 100 <212> DNA
<213> Homo Sapiens 715029 <400> 76 aagctccccc ctactcagtt gccttgaata ggaattatac tgaccaaaar ctagatttaa 60 aatgagaaac tttaacatac ccagttagac ccacctacac 100 <2l0> 77 <211> 100 <212> DNA
<213> Homo Sapiens 719185 <400> 77 tgagaagaga gagcaactgg aagggtgagg ggagatactg agttttagay ggttttcgaa 60 aaagcaagtc actgcccaac aagtgatgtt cagtagacaa 100 <210> 78 <21l> 100 <212> DNA
<213> Homo Sapiens 726176 <400> 78 caaaaaggca gctaaaaaga caaataaata taactggtgg cacttactty agtgtctgag 60 ctttaaagtc aaatttggct ctggcaggtc tcatctgaac 100 <210> 79 <21l> 100 <212> DNA
<213> Homo Sapiens 734463 <400> 79 gttaacatat ctgaaattca taggtaacac ctttcccaag ataatttagy tctggtatta 60 gatttagaat cttcatcctg gtttttttgg tctagtgttt 100 <210> 80 <211> 100 <212> DNA
<213> Homo Sapiens 739485 <400> 80 tcaatagata ggttgatcac tctataaaat agtagtctgt tggcacttay gtgcgtaacc 60 agaggccatc atgttctgag actgccccag ggctttttgg 100 <210> Sl <211> 100 <212 > DNA
<213> Homo Sapiens 765480 <400> 81 agagaagtcc tgcagagggg aaatcataac tcggagtacc tcctaaatay acttgctata 60 atctcacctc cttctatttt ttattttttt ttaattttca 100 <210> 82 <211> 100 <212> DNA
<213> Homo Sapiens 781062 <400> 82 ttttattagt atagtggatg tctcttttgt ttccctttct tgtcatttcw ctttctttcc 60 cctgaggcaa atactattat gaatttagtg tcttcatttt 100 <210> 83 <211> 100 <212> DNA
<213> Homo Sapiens 878201 <400> 83 cttccccgtt gtggactttg ggctgggcct tgcaatttct ctggccaagr gatgttggcc 60 tcctgcctgc tgaggacatg ccctgaagaa cgtgtgcccg 100 <210> 84 <211> 100 <212> DNA
<213> Homo Sapiens 880456 <400> 84 tccattccat tctgtgccat cccatgccac tcccttccct gccactccaw gctattagaa 60 ttggagacag agaccccaat tacttttaat cctcagaatt 100 <210> 85 <211> 100 <212> DNA
<213> Homo Sapiens 886229 <400> 85 ccttaccctt gttagtggat agtggggatt taaaatagca gataatattr tatttagtaa 60 tagttaatat tttttaagga tttattatgt gtgaagtact 100 <210> 86 <2l1> 100 <2l2> DNA
<213> Homo Sapiens 898455 <400> 86 gtctcctttg aacatctggc ctggactcat ccccatgtct gtccatctgy tgtcatccca 60 ggcctggcct gcccacagcc cgggccgcag tcaccactca 100 <210> 87 <211> 100 <212> DNA
<213> Homo sapiens 966557 <400> 87 tgattcacgt gtggcaggga gggccgtcat taatgaaaat gcgagggcgk cgcacgagtg 60 ttgatgactc agcaggcctt tctacttctg tatgagtcag 100 <210> 88 <211> 100 <212> DNA
<213> Homo Sapiens 994174 <400> 88 cttctcatca aataagtttg agtgttaagc atattgtgga atttggagar tggcactgag 60 tactattaga agcagtactg aegcaaatac attgcatact 100 <210> 89 <211> 100 <212> DNA
<213> Homo Sapiens 1009738 <400> 89 tctctccaca gggccccagc caccgaaggc tcaaaaaaaa aaaaaaaaam aaaaaaaaca 60 aaaccctgca tggggagttg tgagaatctg aagcctgata 100 <210> 90 <21l> 100 <212> DNA
<213> Homo Sapiens 1023996 <400> 90 tagagaagaa tattctaaat cttcttgtgt cttctccttg ttggtcttcr ttctaatcat 60 tggaatcatg aagaagttta gtttctcttc tgtgtgactg 200 <210> 91 <211> 100 <212> DNA
<213> Homo Sapiens 1324758 <400> 91 tgaaaaatgc ttcccagaaa tatgctcata tcattcatcc attgacaatm atcatttgtt 60 ggacatcaac tgcttgtact acattagctt atgacatgaa 100 <210> 92 <2I1> 100 <212> DNA
<213> Homo Sapiens 1361107 <400> 92 gtacttattc ccacagccaa cttcaaatga ccccctccct ggcccccccm caagaaaaaa 60 acagggcagc tgaaaagttt gcttaaactt cttaatactt 100 <210> 93 <211> 100 <212> DNA
<213> Homo Sapiens 1367990 <400> 93 ctttaagatc catgtaaaaa tattgcaaat gagggatttt agtagtaagr aaaatttgcc 60 aatttttact cgattagttt ggctattttt atatattctt 100 <210> 94 <211> 100 <212> DNA
<213> Homo Sapiens 1443995 <400> 94 ccatctggcc taccctaggt taacgctagc ttgcattttg ctttgcttty tgattttaat 60 acatatgtca gtattactaa acaatatatt gtttagtttt 100 <210> 95 <211> 100 <212> DNA
<213> Homo Sapiens 1553018 <400> 95 tcaagcgatt cttgtgcctg agcctcttga gtagctggga ttacaggcaw gtgccactgt 60 gcttggctaa ttcatatata tatatatata tatatatata 100 <210> 96 <21l> 100 <212> DNA
<213> Homo Sapiens 1806913 <400> 96 tttaaaatgt tactttattt tagaataagg tagagttgaa aattgtcagr aatacaaaca 60 ttatttaact atttaatcct ttatatatgt gaagttttaa 100 <210> 97 <211> 100 <212> DNA
<213> Homo Sapiens 1806995 <400> 97 ataatactta attttttggc tatgaagagg agaaaaatca tcattttcty atttttctca 60 tcaggggact ctgaagggct tgggaggagt gatctgccca 100 <210> 98 <211> 100 <212> DNA
<213> Homo Sapiens 1966625 <400> 98 atttagtccc agaagggcgt gaggacatca tgtagacaca aaagagcatr caacagcaaa 60 ttattggggg taaatattta atatcttgcg tctggggaac 100 <210> 99 <211> 89 <212> DNA
<213> Homo sapiens 2001433 <400> 99 caccacttta cattctctca acaacagcgc acaagggtwc cagtttctct gcatccttga 60 caacacttgg gattttctgt tattttgat s9 <210> 100 <211> 100 <212> DNA
<213> Homo Sapiens 207.2067 <400> 100 ggttgtgggg ttctccaggt gctctgtcgc ccaggctgga gtgcagtggy gcgatctcgg 60 ctcactgcaa cctcctcctc ccgggtttaa gtgattcttg l00 <210> 101 <211> 100 <212> DNA
<213> Homo sapiens 2012661 <400> 101 aatggttgca tattattgtc aagtgggtgt taccacatga atatttacak ttgttcagct 60 catctgtgaa tttcttctct tcattcccag gactcaccat 100 <210> 102 <211> 100 <212> DNA
<213> Homo Sapiens 2035324 <400> 102 aactaactta gagaagccac taaaggacaa ttcattctaa ttttcacgtr aatcaaagac 60 ttatctgact aatgctgaat aagttgggga gagatttcat 100 <210> 103 <211> 100 <212> DNA
<213> Homo Sapiens 2048326 <400> 103 aaaaagaatg gcatttggca agatgagact ttggttccta actgctctar gacctaggct 60 ggcttgaggt tcaaagtgcc ggttactctc catttgactg 100 <210> 104 <211> 100 <212> DNA
<213> Homo Sapiens 2061254 <400> 104 agcagtggtg agaggggaag agatttttct atgccatgtt tcctgacagk agcagaacct 60 gtcttcagac ctgaaacact tctcctttta tagtcactat 100 <210> 105 <211> 100 <212> DNA
<213> Homo Sapiens 552647 <400> 105 tactgcatga aggaccaaac caacaagcca gtcaggacac caaaagcacm caaggaagca 60 aggaacaaga tgctgcagtt cttaaaataa actgtaggtg 100 <210> 106 <211> 97 <212> DNA
<213> Homo Sapiens 1022573 <400> 106 attgacatca acaaaagtta tcacagggtt aaactgacac acgtttmatg atatgcgtag 60 tcattgctaa gtaagtaggc ataagaacaa ggtaatt 97 <210> l07 <211> 100 <212> DNA
<213> Homo Sapiens 1076245 <400> 107 aggtgccccg gcccagagaa cctgccagtg aaattgcgtc ctacccacar cccctcgttc 60 ccatagctat ggccttaCtt ggggccctCt ttgcctctga 100 <210> 108 <211> 100 <212> DNA
<213> Homo Sapiens 1365259 <400> 108 tattaaccac ctccaacctt ctgtctagag atcatcctta ttatcactaw actctagttt 60 ggctataaag tatataaaga agtagatgtc aaaatggaca 100 <210> 109 <211> 100 <212> DNA
<213> Homo Sapiens 1409133 <400> 109 taagacccct gtcatttgtt caacagatgt tcccctccaa ggctcaccay aatctgcccc 60 aaatccatct ctctggtttt atcttctaaa attccccacc 100 <210> 110 <211> 100 <212> DNA
<213> Homo Sapiens 1566088 <400> 110 gttcctcttc aggtcggtca agaccaggtt tggaactcag cttgttcacr tctgtttcca 60 agtcatcatc ctaaaatgga atagtcatag aagagattac 100 <210> 111 <211> 100 <212> DNA
<213> Homo Sapiens 1934073 <400> 111 tttttcccat tctgtgggtc tgttcacttt gttgtttttg tttttttgtk tgtttttgtt 60 tttgttttgc tatgcataaa ctttttagtt ttatgcaatc 100 <210> 112 <211> 100 <212> DNA
<213> Homo Sapiens 147316 <400> 112 Cttttggggc tgaggtaata gtggaatgot cccttacagg aagctacccr ggacttcaca 60 ggccctgtgc cagtgacttc cttgaaacca ggaatattga 100 <210> 113 <211> 100 <212> DNA
<213> Homo Sapiens 738987 <400> 113 ttgcaaacgt ggagtgaggg aaggggatta aaaagggaga aggtgaatar gtgactttga 60 atccaaccta cagtggttaa ttttcattta actatctggc 100 <210> 114 <211> 100 <212> DNA
<213> Homo Sapiens 740910 <400> 114 attgttaagt ataacagttt gctaagtaag gtgagtggta taatcatatr tttgtaaaaa 60 gcaaaacaaa ataaagccaa aactcgaacc tatctttgtg 100 <210> 115 <211> 100 <212> DNA
<213> Homo Sapiens 764232 <400> 115 tgtcaaaggg gtctttattt aagcctcaga acacatctag tagaagacar gtaatatgag 60 gaaagaggtt gaggtaggac aagactgaaa actttgtaaa 100 <210> 116 <211> 100 <212> DNA
<213> Homo Sapiens 978009 <400> 116 gtccaCaggg CCCCtgCCCt gCCtttCtgt ctgtcataga gcaggaacgr ctgaccatcc 60 agatgtccct cagcgagaaa ccctgactat acaaatccat 100 <210> 117 <211> 100 <212> DNA
<213> Homo Sapiens 1327678 <400> 117 atctccatct ataaaatggg aatcacagta gtacccacct catatgaccr ttgtgaggat 60 tagatgacat tttatatgga aaatgtctgc agtaaaccgt 100 <210> 118 <211> 100 <212> DNA
<213> Homo Sapiens 1954787 <400> 118 ttgagaaatt ttgaggaagt acaaccaaaa gcaattggag actggttaty ggaaggtgcg 60 gaattgggtg aaggcacgat gcctgggtag ctggtgctgc 100 <210> 119 <211> 100 <212> DNA
<213> Homo Sapiens 2049252 <400> 119 acaatctgtc ttgcactaaa actaggcgtt aaagcaccat ctaagaattw tttttttgag 60 ctagaagatg gacccattta agaaaggtca cagttccttc 100 <210> 120 <211> . 100 <212> DNA
<213> Homo Sapiens 154497 <400> 120 agatgagcag ccagttaagg ttaaagaatc agtggatttg gaataaggay ggaggtttgg 60 aggattctgg gcttgggaaa aataagagca aaggtactga 100 <210> 121 <211> 100 <212> DNA
<213> Homo Sapiens 268666 <400> 121 ggtcaggggt tgggctgact catccagacc tctaaacatt atctgaggcy acctaggggg 60 tcagggggag ttaaggcagt ggcctctggc tgcctcggtt 100 <210> 122 <211> 100 <212> DNA
<213> Homo Sapiens 327219 <400> 122 CttCCttCCt tCCttCCttC CttCCttCCt tCCttCCttC CttCCttCCy tCCCtCtCtC 60 tctctttctt tctttctttc ttttattttt gttttgggtg 100 <210> 123 <211> 100 <212> DNA
<213> Homo Sapiens 434363 <400> 123 agagcagaaa cccaaaagct gttagctgaa aaacaaacaa aaaatacctw tggtgaagcg 60 tgatttttgg gtgaagttaa aggcaggaag ctggtagagt 100 <210> 124 <21l> 100 <212> DNA
<213> Homo Sapiens 653220 <400> 124 ccaagtactt acaggatttc tccaaatgtg ttgtattgat aataatctgr acaagctaga 60 aaacaacaaa cacatgtaaa agtcatttaa ctactcaccc 100 <210> 125 <211> 100 <212> DNA
<213> Homo Sapiens 655351 <400> 125 tttcagaagc taaccacctg acttgagact agaattgcac atgttgcatr ccattcttgc 60 ctttgactct ttcagacatt gtatccaata tcatctatca 100 <210> 126 <211> 100 <212> DNA

<213> Homo Sapiens 699863 <400> 126 taaggtacta taattgatat acttgtatac agggtcctta taaaatacar ctcataceta 60 aatccagata tttaataatt gggcccatta agtccaatca 100 <210> 127 <211> 100 <2l2> DNA
<213> Homo Sapiens 878213 <400> 127 gtcatgaaaa ggttaatcta caatgcactt agacagctaa taataagtar aacatttcca 60 aagaaatagg aacctgggat tgtctctttt taataaatta 100 <210> 128 <211> 100 <212> DNA
<213> Homo Sapiens 918880 <400> 128 cccacattag gcataaaaca taaatgtata ccttaataaa tttcggttay gtaaaagagc 60 gcctgggtaa cctcacccag gttgagatgg aaaatgtgcc 100 <210> 129 <211> 100 <212> DI3A
<213> Homo Sapiens 957452 <400> 129 tatacaggat ggggattcca gacaatcaga gtatgatgtt ctcttgtcay tgatcacatt 60 cactagtgac aactaattag cacatccgag taatcaaaaa 100 <210> 130 <211> 100 <212> DNA
<213> Homo Sapiens 1004660 <400> 130 gCdCCgCCtg tCtataCCag tgCtgatCCC aCCtCtCCCC tCCCCCttCy gCCCtCCtgC 60 ctttgcaatt ggctgtttcc ctccagcttt gtccatcctc 100 <210> 131 <211> 100 <2l2> DNA
<213> Homo Sapiens 1152202 <400> 131 gcgggtggag CCCtcctcta gctatcagca tgactgctgt gttttctgay tcaccagttc 60 cctgtccaag gccatcccta agctctgttt tcttccagtt 100 <210> 132 <211> 100 <212> DNA
<2l3> Homo Sapiens 1251587 <400> 132 atggtactat gctgtaataa aaacctgaag tgtgggagaa gtttcagaay caagaactga 60 gaggaagctg aaaggacttt gaggagagtg taagcagaat 100 <210> 133 <211> 100 <212> DNA
<213> Homo Sapiens 1354087 <400> 133 ttttatacaa attcatcttc ttatagttgg gatgttatta caaaagttts gatatcttaa 60 aaatccctaa tgttttagag tttcatattt attgacctga 100 <210> 134 <211> 100 <212> DNA
<213> Homo Sapiens 1545367 <400>~ 134 ctcccttttc aaaggtgtca tttatgtata catattattt ttatttttaw ttcacataca 60 gatacctaaa ctattccatt tttaagagta gaaatcatgc 100 <210> 135 <211> 100 <212> DNA
<213> Homo Sapiens 1549681 <400> 135 aggggagtga acagaggatg aaagagagaa gaaagggcag tctcagctcr ctgtgccaga 60 gtgcttagct cagcgtataa cagggactcc ataaatatgg 100 <210> 136 <211> 100 <212> DNA
<213> Homo Sapiens 1882409 <400> 136 tgcaggcaat ttagtagaca ctgaagcata tttttacctt tccaaattgr aagtggataa 60 caaggtaaca atagcactgt ccctggagcc atcctgtcct l00 <210> 137 <211> 100 <212> DNA
<213> Homo Sapiens 1884688 <400> 137 gatactcatc cctgggtccc acgcccagag agtctgatgt aatcagtgts gggtgtggac 60 agggcattgg gatttttaaa agctccccag gtgattgcag 100 <210> 138 <211> 100 <212> DNA
<213> Homo Sapiens 1932378 <400> 138 ctatatttta cgtaagttca aggtatgaca atgcatatgt ggcaaagatr tatgtagtct 60 gctttataaa aagtgtctaa ggaaacaaaa ttttctttcc 100 <210> I39 <211> 100 <212> DNA
<213> Homo Sapiens 1978355 <400> 139 atgtaacaaa tacaattaaa tggagatgca atttgattta ttgaaaatar catgatttta 60 atgcagatta actgtaggCa gtaccttacc agaatataat 100 <210> 140 <211> 100 <212> DNA
<213> Homo Sapiens 2031262 <400> 140 attctgatag agaccatatg acctccaagt ggattagctt tcaatgaatk ctacaaacca 60 gaagagtctt tcccccatct tcatcttcat tctctaactt 100 <210> 141 <211> 100 <212> DNA
<213> Homo Sapiens 518116 <400> 141 taaggtgaga gaggcagaag gagggtggaa agaatggcgt ttttggtgcr gtgcaaacta 60 atttgcccaa actgaaaaag ctggtgctCt caccattatg 100 <210> 142 <211> 100 <212> DNA
<213> Homo Sapiens 546445 <400> 142 cagtccttcc gccgccgggg ctaacacaat agtgaaagct gcgctcctgr tcgtcggcag 60 gctggcactg ttatacagga aaagggtcgt gtgcgacctt 100 <210> 143 <211> 100 <212> DNA
<213> Homo Sapiens 715005 <400> 143 gcttttaggt ctttaaagca gcacatgatt ttcaagtgag ttattCaaty aatcgtttct 60 gtatcactct gatgcccctt tggactgtct gcttattgaa 100 <210> 144 <211> 100 <212> DNA
<213> Homo Sapiens 763003 <400> 144 ctgagctcag ccagggacca ggaaccaggg accagagtca ggggagccgk ctagaacaga 60 agggcacgtc ggtgctaggc tgggggaggc gcagagatgg 100 <210> 145 <211> 100 <212> DNA
<2I3> Homo Sapiens 913315 <400> 145 atcaggaagg ggaggccaca ggagttacag aggactctgg aaatacagar agatctgtga 60 tttactcagt gggtatttga tagactaaga aggcaaattc 100 <210> 146 <211> 100 <212> DNA
<213> Homo Sapiens 974828 <400> 146 cctgcagttt tccacatgga attcctcttc ttgcactttg gcctaagcay gcattttttc 60 taatggtgga aaatattaaa cagacttttg ctttagaagg 100 <210> 147 <211> 100 <212> DNA
<213> Homo Sapiens 1109525 <400> 147 tcagcccaga gaaagagtca gtcccataat cccagatctc caccacatcy gcacagaggc 60 ctgttaggcc agggggctgc cctgctcccc gctgccccca 100 <210> 148 <211> 100 <212> DNA
<213> Homo Sapiens 1325561 <400> 148 attaaaaggg aactttttat ctcatatttg aaacataagt cacaagaatr actaacattt 60 atgctttaaa ttattaatcc tagttttcag aattttatta 100 <210> 149 <211> 100 <212> DNA
<213> Homo Sapiens 1382794 <400> 149 aaaactttta ccaataattt tcatggagga ttttatgaaa taaccaaaak ctattaaaaa 60 tgcttatgct gaaaaatata aaatattcag ttttctatta 100 <210> 150 <211> 100 <212> DNA
<213> Homo Sapiens 1813081 <400> 150 tgcaaaagag aaggccacac cttgctcaga aagacctgat gtgtggggcy agtgcaaggg 60 gtcaaaaact gaggagaggc cagaaggcca aagcagcttg 100 <210> 151 <211> 100 <212> DNA
<213> Homo Sapiens 1412302 <400> 151 tcacataatg gaagaatggc ccggcctaga ccagaggttc ctaattaggr gtccatgagc 60 cacgttcaat tatatatatt ttttcttctg gtatttttct 100 <210> 152 <211> 100 <212> DNA
<213> Homo Sapiens 1536847 <400> 152 cagtttaagg tggaaagggg gtaggaggaa ggaggaagtg tatttctttm gtcacattag 60 ccagattctc taccttgaca ttgtatctcc cagatctatt 100 <210> 153 <211> 100 <212 > DNA
<213> Homo Sapiens 231727 <400> 153 tctttcctct tttatgtttc cttggctaca tgctgggata ggggctcatr gtaagtttgc 60 cagattcaac caaaaaacgc cacaaaaaca agaaaacagg 100 <210> 154 <211> 100 <212> DNA
<213> Homo Sapiens 235936 <400> 154 atagtatttc ttttcttcag tagaatggtc aggtctctaa gtgctgcagy ggcttgaact 60 cttgatgaca aacatgtagg aaaacttcat gcagccagtt 100 <210> 155 <211> 100 <212> DNA
<213> Homo Sapiens 246103 <400> 155 tttaacctcc ctccctgcgg gcgaataccc agttgtccca gtacattttr ttgatgagtc 60 cttttttgtt tttgtttttt gagacacagt ctcgctttgt 100 <210> 156 <211> 100 <2l2> DNA
<213> Homo Sapiens 260707 <400> 156 cataattgat gcggccttac aaactatttg atgtccaaac cccatcctty accgggccta 60 ccaggcgagg ctagtggagg tgagaagcct cacaatccat 100 <210> 157 <211> l00 <212> DNA
<213> Homo Sapiens 270565 <400> 157 acagtgccac cttcacgcca gggcctggga ggtctgtatg caggttaccr ggaagaataa 60 gtcagtttag tgagcattcg gcccataaaa ccaccgtgca 100 <210> l58 <211> 100 <212> DNA
<213> Homo Sapiens 291030 <400> 158 agtgtgtcac tcctaactct tcttacctcc tttctttggt ttaaggtctm ttagtgtctg 60 tactgggtag tgttttcaca gctttctcct ctgacatggc 100 <210> 159 <211> 100 <212> DNA
<213> Homo Sapiens 386243 <400> 159 cctcacccac gatggcggcc agtctcccga agcgcgaggc cccgcccagy tctcggcctc 60 tcccgcggcc ggtccctgca ccaagctcgc gagctgactg 100 <2I0> 160 <211> 100 <212> DNA
<213> Homo Sapiens 410852 <400> 160 tgggcctggg gggttcatgt ccaatgttgt ccaatcactg gataattctr acatctaact 60 aaacctcttt tataggaaaa agagatgctt taaaattgtt 100 <210> 161 <211> 100 <2l2> DNA
<213> Homo Sapiens 448083 <400> 161 ctgcaggcag actcagatca gcagagaagc atctcgcctc tggctgtgcy gtccaggttg 60 agctgcgtgt ggcagtgagc acagaggaga aatgcaggga 100 <210> 162 <211> 100 <212> DNA
<213> Homo Sapiens 463295 <400> 162 tgggaacttt accacttttt attattttgt tccttttctc tgctggaaay gcgtttgctg 60 tcaatattat gccatatatc gctgtgccaa cttaagttgc 100 <210> 163 <211> 90 <212> DNA
<213> Homo Sapiens 500243 <400> 163 tctctatcag tataacaaag gagtcagact tgatggccts taagggtcct tccaagtact 60 ttaacattct agaaatttac cggacagaag 90 <210> 164 <211> 100 <212> DNA
<213> Homo Sapiens 715029 <400> 164 aagctccccc ctactoagtt gccttgaata ggaattatac tgaccaaaar ctagatttaa 60 aatgagaaac tttaacatac ccagttagac ccacctacac 100 <210> 165 <211> 100 <212> DNA
<213> Homo Sapiens 715491 <400> 165 atgctgtcct agccaatatc tgctcaccct cttcagccat tccctgtacy gcacattcgc 60 tccagccatg actgccatcc acggcccact catcacatac 100 <210> 166 <211> 100 <212> DNA
<213> Homo Sapiens 718830 <400> 166 acatatttga tgacagagac acaatccctg tcctcaagga gctttcatty tagtagagaa 60 gatgaaaacc agtacagttt ggtaagttag atgatattgg 100 <210> 167 <211> 100 <212> DNA
<213> Homo Sapiens 719185 <400> 167 tgagaagaga gagcaactgg aagggtgagg ggagatactg agttttagay ggttttcgaa 60 aaagcaagtc actgcccaac aagtgatgtt cagtagacaa 100 <210> 168 <211> 100 <212> DNA
<213> Homo Sapiens 723053 <400> 168 aagaggaagg gggatgtgaa acatatacag tacattactc tctgcgatcy tcctctgtgc 60 tatcaacaag caagaaaaaa gagatctgga gaaaacagat 100 <210> 169 <211> 100 <212> DNA
<213> Homo Sapiens 725979 <400> 169 gtcactgggc agtccatgac tgattgccgt gctagttgtg tgggcactam tccctttaga 60 tctttagagg aggctgagtt cactttagac agagaggttt 100 <210> 170 <211> 100 <212> DNA
<213> Homo Sapiens 726176 <400> 170 caaaaaggca gctaaaaaga caaataaata taactggtgg cacttactty agtgtctgag 60 ctttaaagtc aaatttggct ctggcaggtc tcatctgaac 100 <210> 171 <211> 100 <212> DNA
<213> Homo Sapiens 726733 <400> 171 atgtttttga gtttgttttg ggaaatctac tttgcacatt ttgttaattr tctttgaaot 60 caagctttcc aoaggttcaa ctgctattct tcccacttag 100 <210> 172 <211> 100 <212> DNA
<213> Homo Sapiens 728552 <400> 172 taggaactga aagtacctga gctgtaaaac tagagaaacc tttcatatgk aaagggcaag 60 ggaaaaagga gttaacatgt ataagcccga ggaagttgaa 100 <210> 173 <211> 100 <212> DNA
<213> Homo Sapiens 728612 <400> 173 agagagctgt actgggagcc tggaaaagga aggtctaatc ggcaatttty ctttctctac 60 ccatgtattt ggaaaaaata atgaattatc tggttgtggg 100 <210> 174 <211> 100 <212> DNA
<213> Homo Sapiens 729531 <400> 174 ccagtaggtg tcagctacgc acagcctttc tctttctaaa tocatatttr taagaaaacg 60 gcctttcatc cataattttc cctagttaaa gggatcaatt 100 <210> 175 <211> 100 <212> DNA
<213> Homo Sapiens 730939 <400> 175 ttattttact tttttgtatt tgaggttttg atcttgctct gttttgtgtr agaaataaat 60 tagaaattca gcattatttc cttctcaaat gtttagccat 100 <210> 176 <211> 100 <212> DNA
<213> Homo Sapiens 734356 <400> 176 cgtgaagcaa aatctcattg gtaagtctga gttgcccaat gagggtttay aagaatacat 60 tcacgttttc taacttcatt accttttctt acttgaattt 100 <210> 177 <211> 100 <212> DNA
<213> Homo Sapiens 734463 <400> 177 gttaacatat ctgaaattca taggtaacac ctttcccaag ataatttagy tctggtatta 60 gatttagaat cttcatcctg gtttttttgg tctagtgttt 100 <210> 178 <21l> 100 <212> DNA
<213> Homo Sapiens 739485 <400> 178 tcaatagata ggttgatcac tctataaaat agtagtctgt tggcacttay gtgcgtaacc 60 agaggccatc atgttctgag actgccccag ggctttttgg 100 <210> 179 <2l1> 100 <212> DNA
<213> Homo Sapiens 742737 <400> 179 ctggtaataa cttggttggt ctgaacccat gcacagctca gaagtgtgar tgagttaata 60 cccccaagga aaacttttga ctaattgggg gaagaaagca 100 <210> 180 <21l> 100 <212> DNA
<213> Homo sapiens 752331 <400> 180 cttccttaat gatcagaaca gagataaaaa tagtgtttag aatgtcctgr aagcctttgg 60 gaatgttaaa ttagaagaca tctattttga tacagatgtg 100 <210> 181 <211> 100 <212> DNA
<213> Homo Sapiens 756827 <400> 181 aaacatttcc ttctgttcca cccctggtat tcttgtatgt gtgttctgcr taggagggcc 60 tgggagttcc tgttgcttct ctgtgaaggg gggaggactt 100 <210> 182 <211> 100 <212> DNA
<213> Homo Sapiens 761057 <400> 182 tcaggttcac atgtgcattt ggttcattat gtgtctttag tgttttaaty tggaacagtt 60 ctcttcccac ctttggatgt aagtggcaca tttgtttccc 100 <210> 183 <211> 100 <212> DNA
<213> Homo Sapiens 763807 <400> 183 aagtaggcaa gaccctaaaa agaagccacc tgCCtCtCCC agtCCCtCCr ggaggactcc 60 aagagggcaa atctgcaggc agtgagctca gggccagcgg 100 <210> 184 <211> 100 <212> DNA
<213> Homo Sapiens 765480 <400> 184 agagaagtcc tgcagagggg aaatcataac tcggagtacc tcctaaatay acttgctata 60 atctcacctc cttctatttt ttattttttt ttaattttca 100 <210> 185 <211> 100 <212> DNA
<213> Homo Sapiens 775251 <400> 185 agctcctttt gatattccct gccccagagc tcatgaccag aacccagagy tgatttaaaa 60 tattttgaaa aatggaggag gcagactgct cccagcagcc 100 <210> 186 <211> 100 <212> DNA
<213> Homo Sapiens 784566 <400> 186 acacaggtcc atctgtggac attgtatatt taaaaaggcc tgtgggaacr atccatgaca 60 gggccactca tagaagaagc tgcaaatttt tagacggtgg 100 <210> 187 <211> 100 <212> DNA

<213> Homo Sapiens 877044 <400> 187 tccctctagc tgctgggggg ctcacttgga tggtcctaaa ggcacccccy ttcttccaat 60 ggtggggtgc cctccagtgt tcacgggatg tgggctcata 100 <210> 188 <211> 100 <212> DNA
<213> Homo Sapiens 878201 <400> 188 cttccccgtt gtggactttg ggctgggcct tgcaatttct ctggccaagr gatgttggcc 60 tcctgcctgc tgaggacatg ccctgaagaa cgtgtgcccg 100 <210> 189 <211> 100 <212> DNA
<213> Homo Sapiens 879496 <400> l89 gctatggttt tttgtttgtt tgtttagcac cagcaggcct gatagcaatr gcaattttta 60 agaaaagaaa aaattagttg ggagaaaggt agtgtatgag 100 <210> 190 <211> 100 <212> DNA
<213> Homo Sapiens 880456 <400> 190 tccattccat tctgtgccat cccatgccac tcccttccct gccactccaw gctattagaa 60 ttggagacag agaccccaat tacttttaat cctcagaatt 100 <210> 191 <211> 100 <212> DNA
<213> Homo Sapiens 885679 <400> 191 caaggaaggg ctattataat tctatttgaa actggctgca tttatcttcw tctctctcag 60 aagtcacatt tcccttgccc ttggatgagg actggaatgc 100 <210> 192 <211> 100 <212> DNA
<213> Homo Sapiens 886229 <400> 192 ccttaccctt gttagtggat agtggggatt taaaatagca gataatattr tatttagtaa 60 tagttaatat tttttaagga tttattatgt gtgaagtact 100 <210> 193 <211> 100 <212> DNA
<213> Homo Sapiens 902336 <400> 193 aatttggaaa ctatagtacg tatcaggtat cgagaaggaa gtgaatataw ctcctataag 60 ctaactttct agtctttttt tcttgagcat ttttttttca 100 <210> 194 <21l> 100 <212> DNA
<213> Homo Sapiens 909802 <400> 194 ttgagtattc ctcagcttta gaagaaaatg tttttctaaa agggaacaay actctgaatt 60 cctttttttc tctaaaatgg aatcactttt tacaaatctc 100 <210> 195 <211> 100 <212> DNA
<213> Homo Sapiens 912142 <400> 195 taatgaacag gatgcacctg ttccatgaac cctctctctc ttccctagar ggagacaccc 60 agatcctcat cagtggcagc agacacctcc aagtctagga 100 <210> 196 <211> 100 <212> DNA
<213> Homo Sapiens 912265 <400> 196 ggctttgggc ataaagcagc cggctggctg ttggatctac agagtcaaay gtgaacaggg 60 aggccttacg aagcaagcct tgttttgctg agagctctgt 100 <210> 197 <211> 100 <212> DNA
<213> Homo sapiens 924799 <400> 197 tctagtgaaa agactaatta ttttttgtat tatcctgttt ccaattttgy ggatcacttc 60 atactaaact gaagacaaac tcctatataa tccttatgtc 100 <210> 198 <211> 100 <212> DNA
<213> Homo Sapiens 975423 <400> 198 tagacttatc ttccaggctc gtgaagagtg cagtaatgct aatgttgagr gatgttaggc 60 acagatcaca tagcatattc ctctcatcat cctgtcagga 100 <210> 199 <211> 100 <212> DNA
<213> Homo Sapiens 985627 <400> 199 tacattgaag accaagccat tctcagtctg ctcttcactt ctaattctam ggttattctg 60 attgctcagg acaatgctgt ctgggaagac tgccccacag 100 <210> 200 <211> 100 <212> DNA
<213> Homo Sapiens 985716 <400> 200 catgaagcag taaatggagc aattaacctg gggcctgaat gggaagaagk ttctggaggc 60 aacagggagt ctggagcaaa atatcaccag atgacaatat 100 <210> 201 <211> 100 <212> DNA
<213> Homo Sapiens 994174 <400> 201 cttctcatca aataagtttg agtgttaagc atattgtgga atttggagar tggcactgag 60 tactattaga agcagtactg acgcaaatac attgcatact 100 <210> 202 <211> 100 <212> DNA
<213> Homo Sapiens 996320 <400> 202 tcctagtttt tgactttact gcctatatta cagaggaagt acggtatgtr ttgaagaaat 60 tagaagtcca cccatagctc aaatatgctt cagtatcctt 100 <210> 203 <211> 100 <212> DNA
<213> Homo Sapiens 997929 <400> 203 tccatgtgcc caacttggaa ttctcttctc tgattttccc ataacttaay tgaagtgaga 60 actttgggga gttcaagcac tgtttcactt atttcctttt 100 <210> 204 <211> 100 <212> DNA
<213> Homo Sapiens 998188 <400> 204 ggcctcttat ggctgtattc tattagctct gggtttactg tgatattatr tatatttgtc 60 ttgggccaaa cagcttgctt gcctcagttg aagtctaaaa 100 <210> 205 <211> 100 <212> DNA
<213> Homo Sapiens 998262 <400> 205 ggcatgaata tatatcaaag tttctttttt cccaaggata aacaaattgy cccagcacca 60 ttaattgatg tactaatgtg ctcaaaacac aaaataaatg 100 <210> 206 <211> 100 <212> DNA
<213> Homo Sapiens 998391 <400> 206 atgcacaaat gaaaggttcc ctgtgctcta tccctgcaca acatccttcr tctgctgtgg 60 gagaaatcat cttaaatact taaatgtaga attctgcagc 100 <210> 207 <211> 100 <212> DNA
<213> Homo Sapiens 1008520 <400> 207 agtctatcta ttcatttatt tttcctccaa acctgcttat tttttttaar taccaatacc 60 tgaaaattgt attttattga actgcatatt aatgcattct 100 <210> 208 <211> 100 <212> DNA
<213> Homo Sapiens 1009127 <400> 208 tgttttacta aagttgattg aaattagtta gttgatattc atcctaccak attttgagtt 60 agagaaattt agatagcaat atttgcacct aaactatttt 100 <210> 209 <211> 100 <212> DNA
<213> Homo Sapiens 1016491 <400> 209 actacaacag aaacacactg tcacacttgc tagaatataa ctgtacttgr gcttctcctt 60 tcctgtgaag tagtgctggg ctttctagag tttaattctc 100 <210> 2l0 <211> 100 <212> DNA
<213> Homo Sapiens 1041111 <400> 210 ctgctgaaag atgaatcact atgaatttgg caggggaatg gaaagatcay gagcagctca 60 gcaatacagt ggccccatct agagatggtc acccaggact 100 <210> 211 <211> 100 <212> DNA
<213> Homo Sapiens 1240709 <400> 211 ggacttccag gcaggatgct gaaggataaa tgtggacacg ggtgagggar gcacctgtgg 60 aatctgcagg gaggctggac tccagtcctg atgtcggccc 100 <210> 212 <21l> 100 <212> DNA
<213> Homo Sapiens 1259291 <400> 212 ggcattatca tctcaaaggg caaaagaagg atggctaaac cattggaaay gtggattcgc 60 tatttttgct atggtcccat ttagccccta cccatgtctg 100 <210> 213 <211> 100 <212> DNA
<213> Homo Sapiens 1324758 <400> 213 tgaaaaatgc ttcccagaaa tatgctcata tcattcatcc attgacaatm atcatttgtt 60 ggacatcaac tgcttgtact acattagctt atgacatgaa 100 <210> 214 <211> I00 <212> DNA
<213> Homo Sapiens 1361107 <400> 214 gtacttattc ccacagccaa cttcaaatga ccccctccct ggcccccccm caagaaaaaa 60 acagggcagc tgaaaagttt gcttaaactt cttaatactt 100 <210> 215 <211> 100 <212> DNA
<213> Homo Sapiens 1367990 <400> 215 ctttaagatc catgtaaaaa tattgcaaat gagggatttt agtagtaagr aaaatttgcc 60 aatttttact cgattagttt ggctattttt atatattctt 100 <2l0> 216 <211> 100 <212> DNA
<213> Homo Sapiens 1395381 <400> 216 ctaaagttcc cttgtcctca gagtgagagt tcatagggct catttcaaar ctctcagcca 60 atgaagaaat taaactggac tgtcccaaag aagatgaaaa 100 <210> 217 <211> 100 <212> DNA
<213> Homo Sapiens 1466386 <400> 217 ctccatgtta ttcctgacaa acagttgtcc catgaacaac atcccaacar atcatacttt 60 tcaagtagca attatcagag ggtgctaata attccaggtt 100 <210> 218 <211> 100 <222> DNA
<213> Homo Sapiens 1473138 <400> 218 gctcttaagt caagtcactt gtggaacttg ccatataaac aatggtttgy aatgtaatgt 60 gttgaagtat aaaatagagt tgtgttcaaa gtactggggt 100 <210> 219 <211> 100 <212> DNA
<213> Homo Sapiens 1541306 <400> 219 acgggcgagt atccatggtg ctgattcttc cagttgtccc aaaacaaaty ttcaatgaat 60 tctgaggccc caaggcttcc cttgttggcc aagtcactga 100 <210> 220 <211> 100 <212> DNA
<213> Homo Sapiens 1541515 <400> 220 cagaagttat ttgctcttgt gcataattga aaacaaatgt tctgtgttgk gtacaattat 60 tttgatctta aaggctatgc ttcgtagaaa aagagaaggt 100 <210> 221 <211> 100 <212> DNA
<213> Homo Sapiens 1544656 <400> 221 cagcttgtcc ccgttaggtg acaataagca gacagaagag ccaagcagcy atggaggaaa 60 atcaggtacc taaatttagc tcaactcctg cctcattcag 100 <210> 222 <2l1> 100 <212> DNA
<213> Homo Sapiens 1618967 <400> 222 ttttagcttc ttttctactc tctcaatctt atccgggctt gtttgggggy gaactcttta 60 tcttttatag aagataaagt ttggttttat taaatttttt 100 <210> 223 <211> 100 <212> DNA
<213> Homo Sapiens 1806913 <400> 223 tttaaaatgt tactttattt tagaataagg tagagttgaa aattgtcagr aatacaaaca 60 ttatttaact atttaatcct ttatatatgt gaagttttaa 100 <210> 224 <211> 100 <212> DNA
<213> Homo Sapiens 1806995 <400> 224 ataatactta attttttggc tatgaagagg agaaaaatca tcattttcty atttttctca 60 tcaggggact ctgaagggct tgggaggagt gatctgccca 100 <210> 225 <211> 100 <212> DNA
<213> Homo Sapiens 1883252 <400> 225 ataaatgtct ggtttaaagg agaaatcaat actgttattc ctctccaaar attaagccca 60 gtctaggaag agaagattgg agttttagac ttgattagtt 100 <210> 226 <211> 100 <212> DNA
<213> Homo Sapiens 1939382 <400> 226 acattaggca agttatatta cttccttgct ccttactatc ttcatctgtm aaatggggaa 60 agtatttaac atagggccaa ctacaagcaa attggctttt 100 <210> 227 <211> 100 <212> DNA
<213> Homo Sapiens 1957979 <400> 227 gaaacagtgg tatatactgc caaaaatgta tcttcctaat tgtcagtcty aaaaaaggta 60 gtttccaaac tacctctttt ttaaacatct atcagttttt 100 <210> 228 <211> 100 <212 > DNA
<213> Homo Sapiens 1964211 <400> 228 tcctagctac tcgggaggct gaggcaggat aatcgtttga accctggagr tggaggttgc 60 agtgagctga gattgtgcca ctgcactcca gcctggacta 100 <210> 229 <211> 100 <212> DNA
<213> Homo Sapiens 1966625 <400> 229 atttagtccc agaagggcgt gaggacatca tgtagacaca aaagagcatr caacagcaaa 60 ttattggggg taaatattta atatcttgcg tctggggaac 100 <210> 230 <211> 100 <212> DNA
<213> Homo Sapiens 2006120 <400> 230 agaaccagag gctgggaagg acaagtcacc tgcttgtgag tggcaggggm aggattggga 60 gcagtggact cggtctcagg actgcggccc tcaggaccaa 100 <210> 231 <211> 100 <212> DNA
<213> Homo sapiens 2012067 <400> 231 ggttgtgggg ttctccaggt gctctgtcgc ccaggctgga gtgcagtggy gcgatctcgg 60 ctcactgcaa cctcctcctc ccgggtttaa gtgattcttg 100 <210> 232 <2I1> 100 <212> DNA
<213> Homo Sapiens 2012661.
<400> 232 aatggttgca tattattgtc aagtgggtgt taccacatga atatttacak ttgttcagct 60 catctgtgaa tttcttctct tcattcccag gactcaccat 100 <210> 233 <211> 100 <212> DNA
<213> Homo Sapiens 2023675 <400> 233 atgcaaatgt gttcatggtc atgggtggta cactggctag ttttaaatay gagccctcag 60 agctgcagga gaaggccaca tgaatgcctc caccacagac 100 <210> 234 <211> 100 <212> DNA
<213> Homo Sapiens 2042417 <400> 234 ggataatgtt atatatacct tgtggatata caccttgtgg tgtgtgtatr tatatatata 60 tatatatata catacacaca caccttgtgg agtgcagtgt 100 <210> 235 <211> 100 <212> DNA
<213> Homo Sapiens 2048326 <400> 235 aaaaagaatg gcatttggca agatgagact ttggttccta actgctctar gacctaggct 60 ggcttgaggt tcaaagtgcc ggttactctc catttgactg 100 <210> 236 <211> 100 <212> DNA
<213> Homo Sapiens 2052832 <400> 236 gggatgatac actctttggt cctaccaaac agctttcatc ttgagatcgy cctccatctt 60 cattgtggcc cttcCtctca gttctctttt gtgtggtccc 100 <210> 237 <211> 100 <212> DNA
<213> Homo Sapiens 2061254 <400> 237 agcagtggtg agaggggaag agatttttct atgccatgtt tcctgacagk agcagaacct 60 gtcttcagac ctgaaacact tctcctttta tagtcactat 100 <210> 238 <211> 100 <212> DNA
<213> Homo Sapiens 2064168 <400> 238 gctggacacc tgggaggggc agcctgtctg tcttctgctc tgccctttcy tgctgtgttc 60 tggaagcata ggcaagcatt ccacagtatc ctgcctgcag 100 <210> 239 <211> 100 <212> DNA
<213> Homo Sapiens 1412302 <400> 239 tcacataatg gaagaatggc ccggcctaga ccagaggttc ctaattaggr gtccatgagc 60 cacgttcaat tatatatatt ttttcttctg gtatttttct 100 <210> 240 <211> 100 <212> DNA
<213> Homo Sapiens 291030 <400> 240 agtgtgtcac tcctaactct tcttacctcc tttctttggt ttaaggtctm ttagtgtctg 60 tactgggtag tgttttcaca gctttctcct ctgacatggc 100 <210> 241 <211> 100 <212> DNA
<213> Homo Sapiens 713970 <400> 241 ccaaacagca ctcagaaaat cataaaatgg caagcaaaat gagcaattcy gcacgcagca 60 cgtgacatgg agtatgagct cgcatttgcc gaaggagcaa 100 <210> 242 <211> 100 <212> DNA
<213> Homo Sapiens 718830 <400> 242 acatatttga tgacagagac acaatccctg tcctcaagga gctttcatty tagtagagaa 60 gatgaaaacc agtacagttt ggtaagttag atgatattgg 100 <210> 243 <211> 100 <212> DNA
<213> Homo Sapiens 723053 <400> 243 aagaggaagg gggatgtgaa acatatacag tacattactc tctgcgatcy tcctctgtgc 60 tatcaacaag caagaaaaaa gagatctgga gaaaacagat 100 <210> 244 <211> l00 <212> DNA
<213> Homo Sapiens 725979 <400> 244 gtcactgggc agtccatgac tgattgccgt gctagttgtg tgggcactam tccctttaga 60 tctttagagg aggctgagtt cactttagac agagaggttt 100 <2l0> 245 <211> 100 <212> DNA
<213> Homo Sapiens 726733 <400> 245 atgtttttga gtttgttttg ggaaatctac tttgcacatt ttgttaattr tctttgaact 60 caagctttcc acaggttcaa ctgctattct tcccacttag 100 <210> 246 <2l1> 100 <212> DNA
<213> Homo Sapiens 728612 <400> 246 agagagctgt actgggagcc tggaaaagga aggtctaatc ggcaatttty ctttctctac 60 Ccatgtattt ggaaaaaata atgaattatc tggttgtggg 100 <210> 247 <211> 100 <212> DNA
<213> Homo Sapiens 734356 <400> 247 cgtgaagcaa aatctcattg gtaagtctga gttgcccaat gagggtttay aagaatacat 60 tcacgttttc taacttcatt accttttctt acttgaattt 100 <210> 248 <211> 100 <212> DNA

<213> Homo Sapiens 742737 <400> 248 ctggtaataa cttggttggt ctgaacccat gcacagctca gaagtgtgar tgagttaata 60 cccccaagga aaacttttga ctaattgggg gaagaaagca 100 <210> 249 <211> 100 <212> DNA
<213> Homo Sapiens 752331 <400> 249 cttccttaat gatcagaaca gagataaaaa tagtgtttag aatgtcctgr aagcctttgg 60 gaatgttaaa ttagaagaca tctattttga tacagatgtg 100 <210> 250 <211> 100 <212> DNA
<213> Homo Sapiens 756827 <400> 250 aaacatttcc ttctgttcca cccctggtat tcttgtatgt gtgttctgcr taggagggcc 60 tgggagttcc tgttgcttct ctgtgaaggg gggaggactt 100 <210> 251 <211> 100 <212> DNA
<213> Homo Sapiens 761057 <400> 251 tcaggttcac atgtgcattt ggttcattat gtgtctttag tgttttaaty tggaacagtt 60 ctcttcccac ctttggatgt aagtggcaca tttgtttccc 100 <210> 252 <211> 100 <212> DNA
<213> Homo Sapiens 879784 <400> 252 gcaggaccgc gcacagagta ggcagcattt gtgtgtgcct cgcacgtgtr tggtggcact 60 gtatgcactt gtagatactc aaatacgtag ggaaagatga 100 <210> 253 <211> 100 <212> DNA
<213> Homo Sapiens 881929 <400> 253 acccagtggt tctgagctct ggcccttcta ctgcctcttg ccagctctgk gtctcagcct 60 tcctatctgt gagttagaca ccaggtagct ggaggggaaa 100 <210> 254 <211> 100 <212> DNA
<2i3> Homo Sapiens 902336 <400> 254 aatttggaaa ctatagtacg tatcaggtat cgagaaggaa gtgaatataw ctcctataag 60 ctaactttct agtctttttt tcttgagcat ttttttttca I00 <210> 255 <211> 100 <212> DNA
<213> Homo Sapiens 912265 <400> 255 ggctttgggc ataaagcagc cggctggctg ttggatctac agagtcaaay gtgaacaggg 60 aggccttacg aagcaagcct tgttttgctg agagctctgt 100 <210> 256 <211> l00 <212> DNA
<213> Homo Sapiens 915941 <400> 256 cggctgcgtc tatggtcatg acgtctgaca gagcgtccac ccgtcttcgm caggactcta 60 tggttcttac gcgcgcagac agaccgccta tataagccat 100 <210> 257 <211> 100 <212> DNA
<213> Homo Sapiens 924799 <400> 257 tctagtgaaa agactaatta ttttttgtat tatcctgttt ccaattttgy ggatcacttc 60 atactaaact gaagacaaac tcctatataa tccttatgtc I00 <210> 258 <2i1> I00 <212> DNA
<213> Homo Sapiens 985627 <400> 258 tacattgaag accaagccat tctcagtctg ctcttcactt ctaattctam ggttattctg 60 attgctcagg acaatgctgt ctgggaagac tgccccacag ' 100 <210> 259 <211> 100 <212> DNA
<213> Homo Sapiens 997929 <400> 259 tccatgtgcc caacttggaa ttctcttctc tgattttccc ataacttaay tgaagtgaga 60 actttgggga gttcaagcac tgtttcactt atttcctttt 100 <210> 260 <21l> 100 <212> DNA
<213> Homo Sapiens 998188 <400> 260 ggcctcttat ggctgtattc tattagctct gggtttactg tgatattatr tatatttgtc 60 ttgggccaaa cagcttgctt gcctcagttg aagtctaaaa 100 <210> 261 <211> 100 <212> DNA
<213> Homo Sapiens 998262 <400> 261 ggcatgaata tatatcaaag tttctttttt cccaaggata aacaaattgy cccagcacca 60 ttaattgatg tactaatgtg ctcaaaacac aaaataaatg 100 <210> 262 <211> 100 <212> DNA
<213> Homo Sapiens 1008520 <400> 262 agtctatcta ttcatttatt tttcctccaa acctgcttat tttttttaar taccaatacc 60 tgaaaattgt attttattga actgcatatt aatgcattct 100 <210> 263 <211> 100 <212> DNA
<213> Homo Sapiens 1009127 <400> 263 tgttttacta aagttgattg aaattagtta gttgatattc atcctaccak attttgagtt 60 agagaaattt agatagcaat atttgcacct aaactatttt 100 <210> 264 <211> 100 <212> DNA
<213> Homo Sapiens 1016491 <400> 264 actacaacag aaacacactg tcacacttgc tagaatataa ctgtacttgr gcttctcctt 60 tcctgtgaag tagtgctggg ctttctagag tttaattctc 100 <210> 265 <211> 100 <212> DNA
<213> Homo Sapiens 1019178 <400> 265 tcaagtgtat tttaccacaa aggcctcttc agggacattt ataacatatm ccagatgtag 60 ttttccaagg agagtgcttt gagaaacact gaccaagcct 100 <210> 266 <211> 100 <212> DNA
<213> Homo Sapiens 1240709 <400> 266 ggacttccag gcaggatgct gaaggataaa tgtggacacg ggtgagggar gcacctgtgg 60 aatctgcagg gaggctggac tccagtcctg atgtcggccc 100 <210> 267 <211> 100 <212> DNA
<213> Homo Sapiens 1375224 <400> 267 aataagcatt ctcaattctg caattcagtg gcattgcata tatcttaggy ggtgtttatg 60 acttcatatt tataatttgg cattcatgtg tttatcttct 200 <210> 268 <211> 100 <212> DNA
<213> Homo Sapiens 1395381 <400> 268 ctaaagttcc cttgtcctca gagtgagagt tcatagggct catttcaaar ctctcagcca 60 atgaagaaat taaactggac tgtcccaaag aagatgaaaa 100 <210> 269 <211> 100 <212> DNA
<213> Homo Sapiens 1455576 <400> 269 acatgaaaga atttttggcc aggcaatgga gatagacagt accattagcy gaaaacaata 60 agacaatgac atgccatact agaaatatta aatcattctt 100 <210> 270 <211> 100 <212> DNA
<213> Homo Sapiens 2008553 <400> 270 tgctctgtat tggctgtggg gtgggggtgg cggtggaact gacccctggs gtctgcattt 60 ctcaggctcc cagggctgtg gctgactttg gccaatggga 100 <210> 271 <211> 84 <212> DNA
<213> Homo Sapiens 242143 <400> 271 ctccaggcga ttcagatgca ggcttaagtc tgagaaccac tgttgtaggy ggttgtggtt 60 ggccccgttc ttagagtaga attc 84 <210> 272 <211> 100 <212> DNA
<213> Homo Sapiens 593930 <400> 272 agcattgctt tgagaataga aatagtcagg aatggttggc tcatgcctgy aatcccaaca 60 ctttgacaag ctgagactat cacttgagac cacgagttca 100 <210> 273 <211> 100 <212> DNA
<213> Homo Sapiens 732148 <400> 273 cctcatgatt tgttctgcca cttttagaat actaaatgct cacttatcty taagatctta 60 tctagagcta tcatattaag tttaggacta ttcctctacc 100 <210> 274 <211> 100 <212> DNA
<213> Homo Sapiens 1377478 <400> 274 ggtaaaatac tgacttgggt cacatgatct acctcaccat agggagaggk atactcactg 60 acctggtcag agccatacac ccaaggagat aacctttgcc 100 <210> 275 <211> 100 <212> DNA
<213> Homo Sapiens 2011741 <400> 275 aagccaatat tctatgttct atttctcagt ttctgtcatc aatagatatm taggcccttc 60 tcccacattc actgaatatc tacagccaac ctatccaata 100 <210> 276 <211> 100 <212> DNA
<213> Homo Sapiens 218855 <400> 276 aaattcatca tgatcttagg tcttctgata tttctaactt caatcctcaw tgctgttatg 60 gattaaatct tttaacccac ttgctattat attgatggga 100 <210> 277 <211> 100 <212> DNA
<213> Homo Sapiens 347867 <400> 277 ttaaacttcc ttccagttag ttttccctgc ctacctcttc cttgctgtcs tttcagactg 60 tgatccctgc ctacaagaga cagtggctgg gattcgggaa 100 <210> 278 <211> 100 <212> DNA
<213> Homo Sapiens 523200 <400> 278 cttgacaagg gtggcagaga ttaaaaaacc ccacgcttta aacaaacatm tttgggggtg 60 gctatggcac cttgaaaaac ccacaaccag cttgacatga 100 <210> 279 <211> 100 <212> DNA
<213> Homo Sapiens 734908 <400> 279 ccctggctaa ggacgccacc ggcctccagg acagggccct gcctcacagk gtgctgcctc 60 cccgctggct ccgcagcaga agcttcagca ggccattcca 100 <210> 280 <211> 100 <212> DNA
<2l3> Homo Sapiens 743221 <400> 280 tccacctata ccatgcttta aaatgagtat ttgggcaggc tgctctagcr tctatgtaca 60 gtgacagcaa aaccttcata ttatgaaggg ggccactcgg 100 <210> 281 <211> 100 <212> DNA
<213> Homo Sapiens 980099 <400> 281 ccttccccca aaacaacata ttatctggtc tgtcatagaa actgaaatty gcatattatt 60 cttccagttg gattagaagc tatgcattaa ttaatatgca 100 <210> 282 <211> 100 <212> DNA
<213> Homo Sapiens 992864 <400> 282 ttcactcttt tgggaaattt acttacctct tagcttttgg ttccaggatr aatatcacta 60 ataggcgtta gtctgctcaa ttaaattcaa taaatatgaa 100 <210> 283 <211> 100 <2l2> DNA
<213> Homo Sapiens 1076556 <400> 283 tagaacagcg tgcacatctg taagcatctg aagactacac tcaagtgccr gatatatagt 60 tattctgaaa acatgtaaca tgtagcaaat gggtttttct 100 <210> 284 <211> 100 <212> DNA
<213> Homo Sapiens 1360554 <400> 284 tctctaaaat caatgaccta aattgagaga atactgatta ttaaaagaar ataatgtcaa 60 caaagatgtc agaaatgtta ttaaggtgaa attgtaacca 100 <210> 285 <211> 100 <212> DNA
<2l3> Homo Sapiens 1530044 <400> 285 ccacatttct cctagttttg tagcaattgc aacacagata tgttaagtgw gtataaagga 60 aaacgttatg gctatttcat cagcaagtgt tttttgccca 100 <210> 286 <2l1> 100 <212 > DNA
<213> Homo Sapiens 1985080 <400> 286 ggagagaatg gagggaaaca aaggatgtga aaaccoagag ggaatgaggr taagtggctc 60 actgtacaat ccacaagtca gaaggtatag ggatgacagg 100 <210> 287 <211> 100 <2I2> DNA
<213> Homo Sapiens 2039055 <400> 287 aatttcttga agtcttggaa atgaaaattt tcagaggtct ttttatgacm ttatagtcct 60 tttggttaaa gcaaaagaat agactgtttt attttgcaaa 100 <210> 288 <211> 100 <212> DNA
<213> Homo Sapiens 327219 <400> 288 CttCCttCCt tCCttCCttC CttCCttCCt tCCttCCttC CttCCttCCy tCCCtCtCtC 60 tctctttctt tctttctttc ttttattttt gttttgggtg 100 <210> 289 <211> 100 <212> DNA
<213> Homo Sapiens 379601 <220>
<221> misc_feature <222> (100)..(100) <223> n is any nucleotide <400> 289 tttttagctg caatcttttc acggaaaaac aatgtaaaag gcataacatr gagaatagag 60 atggaaaaga aagatacaga cttgtaattg cctgacattn 100 <210> 290 <211> 100 <212> DNA
<213> Homo Sapiens 398158 <400> 290 tcacttcttt caagtctctg ttcaaatgtc cccttattag caaggccttm tttgacctcc 60 ctatgtagaa cagtaaaagc ccaccccctt tttcctccta 100 <210> 291 <211> 100 <212> DNA
<213> Homo Sapiens 492933 <400> 291 gtgaggacct ttttactttt tgaaaaatgg caagccaagg acttaacaay tgccaactaa 60 aacacaaatc cttcacttca caaaataaaa ttagatgatc 100 <210> 292 <211> 100 <212> DNA
<213> Homo Sapiens 508118 <400> 292 gtcacagtta agccacagta tcaaactggt gaggcagaaa ggggagagtm atgtaactgg 60 tagtcaagaa acctggattc tatgattaat actaaaacta 100 <210> 293 <211> 100 <212> DNA
<2l3> Homo Sapiens 689594 <400> 293 aataccatca ctttccctac tttgatggca taccttatag tctttcttcy cttcacatgt 60 agtttttctt attagtatgt aagaagcagc ataaggtagt 100 <210> 294 <211> 100 <212> DNA
<213> Homo Sapiens 699863 <400> 294 taaggtacta taattgatat acttgtatac agggtcctta taaaatacar ctcataccta 60 aatccagata tttaataatt gggcccatta agtccaatca 100 <210> 295 <211> 100 <212> DNA
<213> Homo Sapiens 815147 <400> 295 gtgtgtatgc ttttgtatat gtatgtgtat atatatatat atataatttw aaaaatcttc 60 tctatggctt tgataatgct gtttccccat catgtctata 100 <220> 296 <211> 100 <212> DNA
<213> Homo Sapiens 878213 <400> 296 gtcatgaaaa ggttaatcta caatgcactt agacagctaa taataagtar aacatttcca 60 aagaaatagg aacctgggat tgtctctttt taataaatta 100 <210> 297 <211> 100 <212> DNA
<213> Homo Sapiens 914432 <400> 297 atagttcact gctaaattat ctttggtgca aagtagggta caaaacaaar aaagtcaaaa 60 ccacaaatga tcaattccta attatgcaga atagccacca i00 <210> 298 <2I1> 100 <2l2> DNA
<213> Homo Sapiens 946685 <400> 298 gaacccaggc aaggggcatt tgctgactac agtgtgtctt tcagcaaacr cagccggagg 60 ccccccaccc atggaaggga cagccttgcc aagaaaactg 100 <210> 299 <211> 100 <212> DNA
<213> Homo Sapiens 957452 <400> 299 tatacaggat ggggattcca gacaatcaga gtatgatgtt ctcttgtcay tgatcacatt 60 cactagtgac aactaattag cacatccgag taatcaaaaa 100 <2l0> 300 <211> 100 <212> DNA
<213> Homo Sapiens 1004660 <400> 300 gcaccgcctg tctataccag tgctgatccc acctctcccc tcccccttcy gccctcctgc 60 ctttgcaatt ggctgtttcc ctccagcttt gtccatcctc 100 <210> 301 <211> 100 <212> DNA
<213> Homo Sapiens 1011024 <400> 301 agatttggga atcactgaat tgaactgccg gccattggtg gctggtgagr ctgtcctggc 60 agtgatgcag tattctggtc cctgtgccag cactgatggg 100 <210> 302 <211> 100 <212> DNA
<213> Homo Sapiens 1015736 <400> 302 atagtttgac aggaggaacc atccaggctg taattgtggt aaggaatacm tggttgaatt 60 tcacttttta tgtctttatt ttggaattct gtagctaggt 100 <210> 303 <211> 100 <212> DNA
<213> Homo Sapiens 1198718 <400> 303 ataggttatt attctataag aacaagattc agatttagga taaagctcts tgcatataga 60 agctactgaa tggtaaattg ataaagttaa tctttaatat 100 <210> 304 <211> 100 <212> DNA
<213> Homo Sapiens 1377313 <400> 304 aaaagaagat taaatcacat tcagaagaaa acaacatttg ggagaagacr tctagagaca 60 aaaattgtca aaagtttatt gtaaggctat gaaaaatgag 100 <210> 305 <211> 100 <212> DNA
<213> Homo Sapiens 1384866 <400> 305 acattactta aactatcaaa tctttgatga ggtatttctt tgactaagtr taaacacaaa 60 tctcttagac tagaatgcat caatcctcaa ctgtccaccc 100 <210> 306 <211> 100 <212> DNA
<213> Homo Sapiens 1418978 <400> 306 agcagtggca agatcagatt acctcaaggt cacattactg ggcaaagagy tagacttctg 60 cagggccaga agaggtttgg gagctcattt catccattcc 100 <210> 307 <211> 100 <212> DNA
<213> Homo Sapiens 1936999 <400> 307 tttttttttt ttgagatgga gtctcagtct gtcacCCagg ctggagtgcr gtggcatgat 60 cttggctcac tgcaacctct gcatcccagc ttcaagcaat 100 <210> 308 <211> 100 <212> DNA
<213> Homo Sapiens 1955983 <400> 308 caatataaat ccatggcaga aagattttta aaaagaacta tttttataar tataaaataa 60 aacaattaga aaacattgtt ctaacttaat tggcagcatt 100 <210> 309 <2l1> 100 <212> DNA
<213> Homo Sapiens 2031262 <400> 309 attctgatag agaccatatg acctccaagt ggattagctt tcaatgaatk ctacaaacca 60 gaagagtctt tcccccatct tcatcttcat tctctaactt 100 <210> 310 <211> 100 <212> DNA
<213> Homo Sapiens 2047198 <400> 310 tggtagcaaa cctaaagtca aatttttggc tcaattactt caaccaacay agatctaaat 60 agcttatttt ttcttttatt cgggcctgag ctcttattct 100 <210> 311 <211> 100 <212> DNA
<213> Homo Sapiens 2049590 <400> 311 ttatgttaag gacatttatg atatattata tataatatac cataggtcam tagatacaaa 60 aacatatatg tcaaggatat atataacata caatacatgt 100 <210> 312 <211> 100 <212> DNA
<213> Homo Sapiens 2052972 <400> 312 tatatgtaac gacctatttg cagacacaca agtctcaaag aggggtttay atgcagtgac 60 ttttggtaca atccctgtgg ttagtggaag catttatacc 100 <210> 313 <211> 100 <212> DNA
<213> Homo Sapiens 518116 <400> 313 taaggtgaga gaggcagaag gagggtggaa agaatggcgt ttttggtgc~ gtgcaaacta 60 atttgcccaa actgaaaaag ctggtgctct caccattatg 100 <210> 314 <211> 100 <212> DNA

<213> Homo Sapiens 546445 <400> 314 cagtccttcc gccgccgggg ctaacacaat agtgaaagct gcgctcctgr tcgtcggcag 60 gctggcactg ttatacagga aaagggtcgt gtgcgacctt 1.00 <210> 315 <211> 100 <212> DNA
<213> Homo Sapiens 645142 <400> 315 tcaacatatt tggtattcca tggagagttc tcttagtgga gatagttcas gaaggggttt 60 attttcttaa tttgctttct cagttataaa atcttcccaa 100 <210> 316 <211> 100 <212> DNA
<213> Homo Sapiens 715005 <400> 316 gcttttaggt ctttaaagca gcacatgatt ttcaagtgag ttattcaaty aatcgtttct 60 gtatcactct gatgcccctt tggactgtct gcttattgaa 100 <210> 317 <211> 100 <212> DNA
<213> Homo Sapiens 763003 <400> 317 ctgagctcag ccagggacca ggaaccaggg accagagtca ggggagccgk ctagaacaga 60 agggcacgtc ggtgctaggc tgggggaggc gcagagatgg 100 <210> 318 <211> 100 <2I2> DNA
<213> Homo Sapiens 768312 <400> 318 tctgagctgc atcctgaagg ccacagacaa ggcacagaaa gtcactccar aaagagggaa 60 caggatgtac agaatcatgg aacatctact gtgtgccagg 100 <210> 319 <211> 100 <212> DNA
<213> Homo Sapiens 873893 <400> 319 attgaaaaca cctcctggca aagaaggtgt gatttccacg ggacaatgck gatgactttt 60 taattaagat tcttgtcttc agcctgctgc cttaccaggt 100 <210> 320 <211> 100 <212> DNA
<213> Homo Sapiens 913315 <400> 320 atcaggaagg ggaggccaca ggagttacag aggactctgg aaatacagar agatctgtga 60 tttactcagt gggtatttga tagactaaga aggcaaattc 100 <210> 322 <211> 100 <212> DNA
<213> Homo Sapiens 974828 <400> 321 cctgcagttt tccacatgga attcctcttc ttgcactttg gcctaagcay gcattttttc 60 taatggtgga aaatattaaa cagacttttg ctttagaagg 100 <210> 322 <211> 100 <212> DNA
<213> Homo Sapiens 7.107050 <400> 322 tataattacc cttacctaat tttctaatac agtcctaata ttttctagar ccctctccaa 60 aaggtcagta cagactatca ctgcatttgt ttaagtccac 100 <210> 323 <211> 100 <212> DNA
<213> Homo Sapiens 1109525 <400> 323 tcagcccaga gaaagagtca gtcccataat cccagatctc caccacatcy gcacagaggc 60 ctgtt-aggcc agggggctgc cctgctcccc gctgccccca 100 <210> 324 <211> 100 <212> DNA
<213> Homo Sapiens 1325561 <400> 324 attaaaaggg aactttttat ctcatatttg aaacataagt cacaagaatr actaacattt 60 atgctttaaa ttattaatcc tagttttcag aattttatta 100 <210> 325 <211> 100 <212> DNA
<213> Homo Sapiens 1343877 <400> 325 aaaagtgggt ttttggatgc tcttaaatga gaaatcattt cttaccagtw aaagagaaaa 60 aagtaccaag tagagaataa tggggatact acctaagaca 100 <210> 326 <21l> 100 <212> DNA
<213> Homo Sapiens 1382794 <400> 326 aaaactttta ccaataattt tcatggagga ttttatgaaa taaccaaaak ctattaaaaa 60 tgcttatgct gaaaaatata aaatattcag ttttctatta 100 <210> 327 <21l> 100 <212> DNA
<213> Homo Sapiens 1813081 <400> 327 tgcaaaagag aaggccacac cttgctcaga aagacctgat gtgtggggcy agtgcaaggg 60 gtcaaaaact gaggagaggc cagaaggcca aagcagcttg 100 <210> 328 <21l> 100 <212 > DNA
<213> Homo sapiens 1858955 <400> 328 caggetgtgg atgccgaaga gttcattcta cctgcttcat taaataactr atgtgtttct 60 gggcaatcag gcccaatctc agccctgcca cctataatta 100 <210> 329 <211> 100 <212> DNA
<213> Homo Sapiens 2002962 <400> 329 tgatgatgta tgccacgtgc ttgcccctgt ctgcagaagg acagcagcay agtcctgaat 60 tagccaccgg ggccttcaga gaggatcccg ggacagtgtg 100 <210> 330 <211> 100 <212> DNA
<213> Homo Sapiens 2008988 <400> 330 ggttattctt tccaaaattc cgcaaggctc caattaattc catgtgtatk gtatatttaa 60 cacgcatttc ttcctgagaa caacacaaga tggtgtactt 100 <210> 331 <211> 100 <212> DNA
<213> Homo Sapiens 2051772 <400> 331 ttactcttct cagccccagg gagatgcatt ttaccagcag agccacccay acccaagacc 60 agagctatga ccaaggcata tgagctcaaa aagccccagc 100 <210> 332 <211> 54 <212> DNA
<213> Homo sapiens rs5443 <400> 332 catcatctgc ggcatcacgt cygtggcctt ctccctcagt ggccgcctac tact 54 <210> 333 <211> 82 <212> DNA
<213> Homo Sapiens rs4884 <400> 333 gcaggcgcct acttctgggc ggggatcatg tcgtcratgg actggccttt ctccaacttc 60 ttctccattt ccaccatgag ct 82 <210> 334 <211> 60 <212> DNA
<213> Homo Sapiens rs4646 , <400> 334 gcatacctcc tatgggttgt caccaagcta ggtgctattk gccatctgct cctgttcaca 60 <210> 335 <2ll> 59 <212> DNA
<213> Homo Sapiens rs1079598 <400> 335 tgctgtcaga atcacctatt caaaaggcga atcygatcat gtggttcctg ctgccctta 59 <210> 336 <211> 70 <212> DNA
<213> Homo Sapiens rs1800498 <400> 336 tgggggtgtg aagaaaagag ccttgggtty gactagggaa cctggggeca ctccttcctc 60 tgctactaaa 70 <210> 337 <2l1> 95 <212> DNA
<213> Homo Sapiens rs594689 <400> 337 ggattccctc cctttgtagg gtaggagaaa ttggggggac gggcaaggag ytcatttgtg 60 tcccccacct cggggtctct agcccgctgt ctctg 95 <210> 338 <211> 97 <212> DNA
<213> Homo Sapiens rs6003 <400> 338 cctgagtaat ggttacatct ctgatgtaaa gttattgtat aaaattcaag agaacatgcr 60 ttatggttgc gcttcagggt acaaaaccac tggaggg 97 <210> 339 <211> 63 <212> DNA
<213> Homo Sapiens AF100634 <400> 339 gcccagaacc tgatggccct cattagtgct tggctcttay cttggaagca caggcgctga 60 cag 63 c210> 340 <211> 70 c212> DNA
c213> Homo Sapiens rs285 <400> 340 tgcaagggtt ttgcttaatt ctcaattcaa tgtctcttca tcttttagya gctgtggggt 60 tttgttgttg 70 c210> 341 <211> 100 <212> DNA
c213> Homo Sapiens rs2228478 c400> 341 tcatctacgc cttccacagc caggagctcc gcaggacgct caaggaggtg ctgacrtgct 60 cctggtgagc gcggtgcacg cgctttaagt gtgctgggca 100 c210> 342 c211> 7S
c212> DNA
c213> Homo Sapiens rs1800404 c400> 342 gcggtaattt cctgtgcttc tttccagatc gtgcacagaa ctctggcrgc catgctgggt 60 tcccttgcag cactg 75 <210> 343 <211> 62 <212> DNA
<213> Homo Sapiens rs2065160 <400> 343 gcctctcgat gagtaaatat gggycaaatg aagtaaacag actatgcgta ggtatcatca 60 as 62 <210> 344 <211> 69 <212> DNA
<213> Homo Sapiens rs1042602 <400> 344 ttatgtgtca atggatgcac tgcttggggg atmtgaaatc tggagagaca ttgattttgc 60 ccatgaagc 69 <210> 345 <211> 79 <212> DNA
<213> Homo Sapiens rs3287 <400> 345 ccatcctcca acacacacac aaataagcag ctaatggcaa trctagtggt cttcccaatt 60 cacaagacct gtgcttcaa 79 <210> 346 <211> 52 <212> DNA
<213> Homo Sapiens rs17203 <400> 346 cttcaaattg ctttaagtac tctasgaaga ccttgactgt tggatttttg ag 52 <210> 347 <211> 69 <212> DNA
<213> Homo Sapiens rs3340 <400> 347 ccatttcttt gtaaaataac aataactttr aaggcaaaag caagattctg taaaccaaca 60 ttggaaaag <210> 348 <211> 66 <212> DNA

<213> Homo Sapiens rs2763 <400> 348 gtgaccctgt gaggtcaggt cccccagatt gasgtctgag tgtgggcaag tgtgtcaaaa 60 ggggct 66 <210> 349 <211> 62 <212> DNA
<213> Homo Sapiens rs2161 <400> 349 caggccaaga gcgtcctatg aatoatrcat ttgttcctgt tattgctgtt cacagagtgg 60 ca 62 <210> 350 <211> 60 <212> DNA
<213> Homo Sapiens rs2695 <400> 350 atttgtgttg ggtggtctag rctattcaga aaatctcaga ggaggacaaa tgatagtgca 60 <210> 35l <211> 84 <212> DNA
<213> Homo Sapiens rs2862 <400> 351 catctgagtg caagataaaa aggaaatagc aattcaaggc acaaagctaa gyacatgcaa 60 caatatagat gatttggggg tggg 84 <210> 352 <211> 67 <212> DNA
<213> Homo Sapiens rs2891 <400> 352 aaaaaggcag gacattccaa ggctctctaa caygagtgtc tgcagcccca ttcgctttga 60 gatgtga <210> 353 <211> 80 <212> DNA
<213> Homo Sapiens rs2816 <400> 353 ctccttggca gggatttgtg acactgcatt gctgggctgt gttccycggg ctcttctgga 60 ccttgcaccg tggataccag 80 <210> 354 <211> 66 <212> DNA
<213> Homo Sapiens rs3309 <400> 354 gcctgagaat cccacacatt tcagtccaag awaaccttcc tcaaattttc catctcccat 60 cagagg 66 <210> 355 <211> 60 <212> DNA
<213> Homo Sapiens rs33I7 <400> 355 tccattcagt gtatcacatc ttcaggatag gtrataacag tgtgaagggt gtgctcattt 60 <210> 356 <2II> 56 <212> DNA
<213> Homo Sapiens rs140864 <220>
<221> misc_feature <222> (29) .(29) <223> n is TTC or absent <400> 356 caaaatctgc tccatgtcca atctgctcna agttcaacca acctccctaa cccttg 56 <210> 357 <211> 47 <212> DNA
<213> Homo sapiens rs16626 <220>
<221> misc feature <222> (28)J.(28) <223> n is CGTGAAGTCC or absent <400> 357 atattcacct Ccagccacca gtcacccnta tgggttggga cctctgg 47 <210> 358 <211> 63 <212> DNA
<213> Homo Sapiens rs16434 <220>
<221> misc_feature <222> (41) .(41) <223> n is TCCCACGCGAGTGTGGTGGGACCTTG or absent <400> 358 tctgcttctc ttgagataca tagttaaagt agacacggcc nagacttagg aggattgcct 60 gtt 63 <210> 359 <211> 465 <212> DNA
<213> Homo Sapiens X00238 <220>
<221> misc_feature <222> (152)..(152) <223> n is AATTACAGGTAGAGGGCTAGAAGTTTTTGGACATTAACTATTTCTATCTTCTGATTTAGT
TAACGAGAAACAAAAAATCCTGCAGACAAGTTTCTCCTCAGTCAGGTA or TGGGTATGAACCA
AGTTTGTTTCCTTGGTTAG
<400> 359 ccacaggtgt aacattgtgt tttccttgtc tgtgccaggg acaccttggc atcagatgcc 60 tgaaggtagc agcttgtccc tctttgcctt ctctaattag atatttctct ctctctctcc 120 ctctctccataaagaaaactatgagagagggntttcctaaccaagtttgagggtatgaac 180 atactctccttttccttttctataaagctgaggagaagagtgagggagtgtgggcaagag 240 aggtggctcaggctttccctgggcctgattgaactttaaaacttctctactaattaaaca 300 acactgggctctacactttgcttaaccctgggaactggtcatcagcctttgacctcagtt 360 ccccctcctgaccagctctctgccccaccctgtcctctggaacctCtgcgagatttagag 420 gaaagaacca gttttcaggc ggattgcctc agatcacact atctc 465 ~210> 360 <21l> 97 a212> DNA
<213> Homo sapiens X53550 <220>
<221> misc feature <222> (58)~. . (58) <223> n is (Alu) or absent <400> 360 aagtgctgta ggccatttag attagagaag aaaagagcct tctcaagaag gccttctnaa 60 aactcctctg agtctgtata cgctgtcatc gaagact 97 <210> 361 <211> 141 <212> DNA
<213> Homo Sapiens AL392048 <400> 361 cccgggtgtg cgtagggcgg gcgccaaggc ggctcggcgg ggatccgtcc tcgccagggg 60 ccggrtcccg gcgggaggag gcgccctccc tgccccccgc cacggcggag cgtctgcaga 120 atggtgacag gattctgggt t 141 <210> 362 <211> 127 <212> DNA
<213> Homo Sapiens M57427 <220>
<221> misc_feature <222> (77) .(77) <223> n is (Alu) or absent <400> 362 aactgggaaa atttgaagag aaagtacaca gatacatttg taaggttgtc tgttacttag 60 gctacagagg aaagaanaga attccctctc tctaaacaca ctctaacaca caggagttga 120 gaactca 127 <210> 363 <211> 136 <212> DNA
<213> Homo Sapiens AF015160 <220>
<221> misc_feature <222> (106)..(106) <223> n is (Alu) or absent <400> 363 tctaggccca gatttatggt aactgataat tattgataat ttctttagat ttgttgatat 60 tctacttgat gaaacatcca aattttactt tatgtaaatg tggccnaaaa gtgtcagaaa 120 ataaccaatt gtgctt 136

Claims

What is claimed is:

1. A method of inferring, with a predetermined level of confidence, a trait of an individual, comprising:
a) contacting a sample comprising nucleic acid molecules of a test individual with hybridizing oligonucleotides, wherein the hybridizing nucleotides can detect nucleotide occurrences of single nucleotide polymorphisms (SNPs) of a panel of at least about ten ancestry informative markers (AIMs) indicative of a population structure correlated with the trait, and wherein said contacting is performed under conditions suitable for detecting the nucleotide occurrences of the AIMs of the test individual by the hybridizing oligonucleotides; and b) identifying, with a predetermined level of confidence, a population structure that correlates with the nucleotide occurrences of the AIMs in the test individual, wherein the population structure correlates with a trait, thereby inferring, with a predetermined level of confidence, the trait of the individual.

2. The method of claim 1, wherein the panel comprises at least about twenty AIMs.

3. The method of claim 1, wherein the trait comprises biogeographical ancestry (BGA).

4. The method of claim 3, wherein the panel comprises AIMs set forth as SEQ ID
NOS:1 to 331.

5. The method of claim 3, wherein the panel comprises AIMs set forth as SEQ ID
NOS:1 to 71.

6. The method of claim 3, wherein the panel comprises AIMs set forth as:
SEQ ID NOS:7, 21, 23, 27, 45, 54, 59, 63, and 72 to 152;
SEQ ID NOS:3, 8, 9, 11, 12, 33, 40, 59, 63, and 153 to 239;
SEQ ID NOS:1, 8, 11, 21, 24, 40, 172, and 240 to 331; or combinations thereof.

7. The method of claim 1, wherein at least one AIM of the panel is not linked to a gene linked to the trait.

8. The method of claim 3, wherein the BGA comprises a proportion of a sub-Saharan African, Native American, IndoEuropean, or East Asian ancestral group, or a combination of said ancestral groups.

9. The method of claim 8, wherein the BGA comprises a proportion of at least three ancestral groups.

10. The method of claim 3, wherein the BGA comprises proportions of at least sub-Saharan African and IndoEuropean ancestral groups; Native American and IndoEuropean ancestral groups; East Asian and Native American ancestral groups; or IndoEuropean and East Asian ancestral groups.

11. The method of claim 3, wherein the BGA comprises proportions of at least Native American, East Asian, and IndoEuropean ancestral groups; or sub-Saharan African, Native American, and IndoEuropean ancestral groups.

12. The method of claim 1, wherein the trait comprises responsiveness of the individual to a drug.

13. The method of claim 12, wherein the drug is a cancer chemotherapeutic agent.

14. The method of claim 12, wherein the drug is a statin.

15. The method of claim 1, wherein the trait comprises susceptibility to a disease.

16. The method of claim 15, wherein the disease has an ethnic predisposition.

17. The method of claim 16, wherein the disease is a cancer, diabetes, or hypertension.

18. The method of claim 17, wherein the cancer is prostate cancer.

19. The method of claim 15, wherein the disease is a neurological disorder.

20. The method of claim 19, wherein the schizophrenia or Parkinson's disease.

21. The method of claim 15, wherein the disease is alcoholism.

22. The method of claim 1, wherein the trait comprises a pigmentation trait.

23. The method of claim 22, wherein the pigmentation trait comprises eye color, shin color, hair color, or a combination thereof.

24. The method of claim 1, further comprising identifying, with a predetermined level of confidence, a sub-population structure of the population structure that correlates with the nucleotide occurrences of the AIMs in the test individual, wherein the sub-population structure correlates With a trait.

25. The method of claim 1, wherein the hybridizing oligonucleotides comprise oligonucleotide primers, said method further comprising contacting the sample with a polymerase, under condition suitable for generation of a primer extension product, wherein determining the nucleotide occurrence of a SNP comprises detecting the presence of the primer extension product.

26. The method of claim 1, wherein the hybridizing oligonucleotides comprise oligonucleotide primers, said method further comprising contacting the sample with a polymerase, under condition suitable for generation of a primer extension product, wherein determining the nucleotide occurrence of a SNP comprises determining the nucleotide sequence of the primer extension product at a position corresponding to the position of the SNP.

27. The method of claim 1, wherein the hybridizing oligonucleotides comprise amplification primer pairs, said method further comprising contacting the sample with a polymerase, under condition suitable for generation of an amplification product, wherein determining the nucleotide occurrence of a SNP comprises detecting the presence of the amplification product.

28. The method of claim 1, wherein the hybridizing oligonucleotides comprise amplification primer pairs, said method further comprising contacting the sample with a polymerase, under condition suitable for generation of an amplification product, wherein determining the nucleotide occurrence of a SNP comprises determining the nucleotide sequence of the amplification product at a position corresponding to the position of the SNP.

29. The method of claim 1, which is performed in a high throughput format.

30. The method of claim 1, which is performed in a multiplex format.

31. A method of estimating, with a predetermined level of confidence, proportional ancestry of at least two ancestral groups of an individual, comprising:
a) contacting a sample comprising nucleic acid molecules of a test individual with hybridizing oligonucleotides, wherein the hybridizing oligonucleotides can detect nucleotide occurrences of single nucleotide polymorphisms (SNPs) of a panel of at least about ten ancestry informative markers (AIMs) indicative of biogeographical ancestry (BGA) for each ancestral group examined, and wherein said contacting is under conditions suitable for detecting the nucleotide occurrences of the AIMs of the test individual by the hybridizing oligonucleotides; and b) identifying, with a predetermined level of confidence, a population structure that correlates with the nucleotide occurrences of the AIMs of each of the ancestral groups examined, wherein the population structure is indicative of proportional ancestry, thereby estimating, with a predetermined level of confidence, proportional ancestry of the individual.

32. The method of claim 31, wherein the proportional ancestry comprises a proportion of a sub-Saharan African ancestral group, a Native American ancestral group, an IndoEuropean ancestral group, an East Asian ancestral group, or a combination thereof.

33. The method of claim 31, wherein the proportional ancestry comprises proportions of sub-Saharan African and IndoEuropean ancestral groups; Native American and IndoEuropean ancestral groups; East Asian and Native American ancestral groups; or IndoEuropean and East Asian ancestral groups.

34. The method of claim 31, wherein the proportional ancestry comprises proportions of Native American, East Asian, and IndoEuropean ancestral groups; sub-Saharan African, Native American, and IndoEuropean ancestral groups; or sub-Saharan African, Native American, and East Asian ancestral groups.

35. The method of claim 31, wherein the panel for at least one of the ancestral groups comprises AIMs as set forth in SEQ ID NOS:1 to 331.

36. The method of claim 31, wherein the panel for at least one of the ancestral groups comprises AIMs set forth as SEQ ID NOS:1 to 71.

37. The method of claim 31, wherein the panel for at least one of the ancestral groups comprises AIMs set forth as:
SEQ ID NOS:7, 21, 23, 27, 45, 54, 59, 63, and 72 to 152;
SEQ ID NOS:3, 8, 9, 11, 12, 33, 40, 59, 63, and 153 to 239;
SEQ ID NOS:1, 8, 11, 21, 24, 40, 172, and 240 to 331; or combinations thereof.

38. The method of claim 31, wherein at least one AIM of the panel is not linked to a gene linked to the trait.

39. The method of claim 31, wherein the proportional ancestry comprises proportions of three ancestral groups, and wherein identifying a population structure that correlates with the nucleotide occurrences of the AIMs of the test individual comprises:
performing a likelihood determination for affiliation with each of a sub-Saharan African ancestral group, a Native American ancestral group, an IndoEuropean ancestral group, and an East Asian ancestral group;
thereafter selecting three ancestral groups having a greatest likelihood value;

determining a likelihood of all possible proportional affiliations among the three ancestral groups having the greatest likelihood value, whereby a population structure or proportional affiliation that correlates with the nucleotide occurrences of the AIMs of the test individual is identified; and identifying a single proportional combination of maximum likelihood, thereby estimating the proportional ancestry of the individual.

40. The method of claim 31, wherein the proportional ancestry comprises proportions of three ancestral groups, and wherein identifying a population structure that correlates with the nucleotide occurrences of the AIMs comprises:
performing six two-way comparisons comprising likelihood determinations for affiliation between each group with each other group;
thereafter selecting three ancestral groups having a greatest likelihood value;
determining a likelihood of all possible proportional affiliations among the three ancestral groups having the greatest likelihood value, whereby a population structure or proportional affiliation that correlates with the nucleotide occurrences of the AIMs of the test individual is identified; and identifying a single proportional combination of maximum likelihood, thereby estimating the proportional ancestry of the individual.

41. The method of claim 31, wherein the proportional ancestry comprises proportions of three ancestral groups, and wherein identifying a population structure that correlates with the nucleotide occurrences of the AIMs of the test individual comprises performing three three-way comparisons among the groups, determining a likelihood of all possible proportional affiliations among the three ancestral groups having the greatest likelihood value, whereby a population structure or proportional affiliation that correlates with the nucleotide occurrences of the AIMs of the test individual is identified; and identifying a single proportional combination of maximum likelihood, thereby estimating the proportional ancestry of the individual.

42. The method of claim 31, wherein the proportional ancestry comprises proportions of four ancestral groups, and wherein identifying a population structure that correlates with the nucleotide occurrences of the AIMs of the test individual comprises performing six two-way comparisons, three three-way comparisons, or one four-way comparison among the groups;
determining a likelihood of all possible proportional affiliations among the four ancestral groups having the greatest likelihood value, whereby a population structure or proportional affiliation that correlates with the nucleotide occurrences of the AIMs of the test individual is identified; and identifying a single proportional combination of maximum likelihood, thereby estimating the proportional ancestry of the individual.

43. The method of claim 40, further comprising generating a graphical representation of the comparison of the three ancestral groups, said graphical representation comprising a triangle with each an ancestral group independently represented by a vertex of the triangle, wherein the maximum likelihood value of proportional affiliation for an individual comprises a point within the triangle.

44. The method of claim 43, wherein the graphical representation further comprises a confidence contour indicating a level of confidence associated with estimating the proportional ancestry.

45. The method of claim 31, further comprising identifying, with a predetermined level of confidence, a sub-population structure of the population structure that correlates with the nucleotide occurrences of the AIMs in the test individual, wherein the sub-population structure correlates with ethnicity of the test individual.

46. The method of claim 45, wherein identifying the subpopulation structure comprises identifying chromosomes of the test individual that contain the AIMs indicative of an ancestral group of the proportional ancestry of the test individual, contacting a sample comprising nucleic acid molecules of the test individual with second hybridizing oligonucleotides;
wherein the second hybridizing oligonucleotides can detect nucleotide occurrences of SNPs of a second panel of AIMs, and wherein the AIMs of the second panel are present on the chromosomes of the test individual that contain the AIMs indicative of the ancestral group of the test individual; and identifying a sub-population structure that correlates with the nucleotide occurrences of the AIMs of the second panel, wherein the sub-population is indicative of ethnicity of the ancestral group of the test individual.

47. The method of claim 45, wherein the ancestral group is IndoEuropean, and wherein the ethnicity comprises Northern European or Mediterranean.

48. The method of claim 31, further comprising generating an ancestral map of the world, wherein locations of populations having a proportional ancestry corresponding to the proportional ancestry of the test individual are indicated on the ancestral map.

49. The method of claim 48, further comprising a) overlaying the ancestral map with a genealogical map, wherein the genealogical map indicates locations of populations having geopolitical relevance with respect to the test individual, and b) statistically combining the information of the ancestral map and genealogical map to obtain a most likely estimate of family history of the test individual.

50. The method of claim 31, wherein identifying a population structure that correlates with the nucleotide occurrences of the AIMS comprises comparing the nucleotide occurrences of the AIMS of the test individual with known proportional ancestries corresponding to nucleotide occurrences of AIMS indicative of BGA.

51. The method of claim 50, wherein the known proportional ancestries corresponding to nucleotide occurrences of AIMS indicative of BGA are contained in a database.

52. The method of claim 51, wherein the comparing is performed using a computer.

53. The method of claim 50, wherein each of the known proportional ancestries corresponding to nucleotide occurrences of AIMS indicative of BGA further comprises of photograph of a person from whom the known proportional ancestry was determined.

54. The method of claim 53, wherein the photograph comprises a digital photograph.

55. The method of claim 54, wherein digital information comprising the digital photograph is contained in a database.

56. The method of claim 55, wherein the digital information in the database is associated with a known proportional ancestry corresponding to nucleotide occurrences of AIMS indicative of BGA of the person in the photograph.

57. The method of claim 51, further comprising identifying a photograph of a person having a proportional ancestry corresponding to the proportional ancestry of the test individual.

58. The method of claim 57, wherein identifying the photograph comprises scanning a database comprising a plurality of files, each file comprising digital information corresponding to a digital photograph of a person having a known proportional ancestry, and identifying at least one photograph of a person having nucleotide occurrences of AIMS
indicative of BGA that correspond to the nucleotide occurrences of AIMs indicative of BGA
of the test individual.

59. An article of manufacture, comprising at least one photograph of a person having a known proportional ancestry corresponding to a population structure comprising nucleotide occurrences of AIMs indicative of biogeographical ancestry (BGA).

60. The article of claim 59, which is contained in a file.

61. A plurality of files comprising the article of manufacture of claim 59, wherein files of the plurality comprise at least one photograph of a person having a known proportional ancestry corresponding to a population structure comprising nucleotide occurrences of AIMS indicative of BGA.

62. The file of claim 60, which comprises a plurality of photographs, wherein photographs of the plurality comprise a photograph of a person having a known proportional ancestry corresponding to a population structure comprising nucleotide occurrences of AIMS
indicative of BGA.

63. The file of claim 62, wherein photographs of the plurality comprise photographs of different persons having the same known proportional ancestries.

64. The file of claim 62, wherein photographs of the plurality comprise photographs of different persons having different known proportional ancestries.

65. The article of manufacture of claim 59, wherein the at least one photograph comprises a digital photograph.

66. The article of manufacture of claim 65, wherein the digital photograph comprises digital information.

67. The article of manufacture of claim 66, wherein the digital information is contained in a database.

68. The article of manufacture of claim 65, comprising a plurality of digital photographs.

69. A plurality of the articles of manufactures of claim 65, comprising at least two digital photographs.

70. The plurality of claim 69, wherein the digital photographs comprise digital information.

71. The plurality of claim 70, wherein the digital information is contained in a database.

72. A kit, comprising a plurality of hybridizing oligonucleotides, which comprise at least fifteen contiguous nucleotides of at least five polynucleotides as set forth in SEQ ID
NOS:1 to 331, or polynucleotides complementary thereto.

73. The kit of claim 72, wherein the hybridizing oligonucleotides comprise at least fifteen contiguous nucleotides of at least five polynucleotides as set forth in SEQ ID NOS:1 to 71, or polynucleotides complementary thereto.

74. The kit of claim 72, wherein hybridizing oligonucleotides of the plurality comprise at least one nucleotide corresponding to a position of the polymorphism of the polynucleotide, or polynucleotide complementary thereto.

75. The kit of claim 74, wherein hybridizing oligonucleotides of the plurality comprise nucleotide position 50 of a polynucleotide as set forth in any of SEQ
ID NOS:1 to 34, 36 to 49, 52 to 55, or 57 to 98, 100 to 105, 107 to 162, 164 to 331, or a polynucleotide complementary thereto.

76. The kit of claim 72, wherein hybridizing oligonucleotides of the plurality comprise at least one probe, at least one primer, or a combination thereof.

77. The kit of claim 76, comprising at least one amplification primer.

78. The kit of claim 76, comprising at least one amplification primer pair, which comprises a forward primer and a reverse primer.

79. The kit of claim 78, further comprising reagents for performing an amplification reaction using the at least one amplification primer pair.

80. The lit of claim 72, further comprising at least one ancestry informative marker (AIM), wherein the AIM corresponds to a hybridizing oligonucleotide of the plurality.

81. The kit of claim 72, further comprising a detectable label, which can be bound to or incorporated into at least one hybridizing oligonucleotide of the plurality.

82. The kit of claim 72, wherein hybridizing oligonucleotides of the plurality are detestably labeled.