JP2012050432A - Composition and method for inferring ancestry - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide ancestry informative markers (AIMs) containing a single nucleotide polymorphism, and to provide a method for using a panel of AIMs for drawing an inference as to a trait of an individual.SOLUTION: The method for inferring the trait of the individual includes following steps of: (a) contacting a nucleic acid containing sample of a test individual with an oligonucleotide probe that can detect nucleotide occurrences of single nucleotide polymorphisms (SNPs) of a panel of at least about ten AIMs wherein the population structure correlates with the trait; 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 the trait, and inferring, with a predetermined level of confidence, the trait of the individual, which are biogeographical ancestry, pigmentation traits, drug responsiveness, or disease susceptibility of the individual.

Description

Field of the Invention The present invention is generally useful as identification of genetic markers that predict an individual's biogeographic ancestry, and more particularly as an ancestor information marker (AIM) that allows inferences about an individual's traits To infer individual single nucleotide polymorphism combinations, algorithms for identifying such AIMs, and individual traits, including individual ancestry, individual responsiveness to drugs, and individual predisposition for disease It relates to a method that uses AIM.

Background information The majority of genetic variation (80-90%) among human individuals is between individuals, and only a relatively small percentage (10-20%) is due to population differences (Nei, Molecular Population Genetics ( Columbia University Press, New York) 1987; Cavalli-Sforza et al., The History and Geography of Human Genes (Princeton University Press, Princeton, NJ) 1994; Deka et al., Electrophoresis 16: 1659-1664, 1995; Rosenberg et al., Science 298: 2381 -2385, 2002; Akey et al., BioTechniques 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. Very few classical markers (e.g. blood group, serum proteins and immunological markers) or DNA genetic markers that are population-specific or have large frequency differences between geographically and ethnically defined populations ( Roychodhury and Nei, Human Polymorphic Genes: World Distribution (Oxford University Press, New York) 1988; Dean et al., Amer. 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, significant physical fitness among human populations, probably reflecting inherent ecological conditions, accidental genetic drift and genetic adaptation to gender selection And there are physiological differences. In modern populations, these differences are evident in morphological differences between ethnic groups, as well as differences in drug responsiveness and disease susceptibility and tolerance.

  At a basic level, the human population structure is a “geographic” genetic component or heritage, and is associated with any measure of determination, BioGeographical Ancestry ( BGA). For example, at a coarse level, BGA can be determined for two groups (eg, European vs. others); or at a fine level, for example, it is Indian Europeans, East Asians, Sub-Saharan Africans and Native Americans “Race” can be referred to by four groups such as (Native American); or at a fine level, for example, it can refer to ethnicity within the European group (eg Mediterranean people) Or Scandinavian); or at a finer level, for example, it extends to groups of families within ethnic groups, such as a group of O'Reilly descendants from a set of common ancestry within the Irish group Can also be mentioned. BGA measurements are relevant for almost any type of genetic or epidemiological study design. For example, BGA is an important component in drug response variability (Burroughs et al., J. Natl. Med. Assoc. 94: 1-26, 2002). The reason for this relationship is that genetic drift, geographic and / or reproductive isolation, and local selective pressure are compatible with alkaloids, tannins (self-defending chemicals), and other xenobiotics found in native foods It has formed the allelic frequency of our ancestors about sex. Most drugs are derived from such chemicals, and therefore it is not a coincidence that the family of enzymes that allow humans to detoxify drugs is found at different frequencies in different populations . This scenario is not specific to drug responsiveness, and many other parts of the genome that are not associated with drug responsiveness are subject to these same types of pressure.

  Researchers generally identify genetic variants that cause disease (so-called “phenotypically active” loci), rather than identifying genetic variants that are simply correlated with the disease. I have been involved. As such, no matter what trait is examined, and for most research designs involving unrelated individuals, trait values in a given sample rather than those in linkage disequilibrium (LD) with a phenotypically active locus. In order to avoid identifying markers of correlated structure, it has been considered important to control the population structure (Risch et al., Genome Biology 3: 1-12, 2001; Wang et al., Amer. J. Hum. Genet. 71: 1227-1234, 2002; Burroughs et al., Supra, 2002; Rao and Chakraborty, Amer. J. Hum. Genet. 26: 444-453, 1974). There are two sources of population structure in the sample collection: 1) sampling effects that can produce structure, even when sampling is done from a homogeneous population, and 2) natural human demographics. The primary source of population structure is awkward to genetics research, and the associations found from this type of structural research are generally an artifact of the collection process rather than reflecting human demographics. Is considered. Most geneticists generally also view the second type of structure as a nuisance. As such, associations identified as due to population structure have been considered false findings or artifacts and have generally been rejected; only true linkage or LD findings have been published, but such markers Because it is considered to be linked to a biologically relevant gene.

  Much effort has been directed to quantify both types of population structure (above) in groups of individuals. Such methods inherently measure deviations from the expected level of heterozygosity within a group of samples as structural indicators (although none of these methods can read intra-individual structures). ). Many common diseases show loci and / or allelic heterogeneity as a function of BGA, and many authors have so far gained improper attention to population structure during the study design phase Suggested that it produced at least some of the so-called “false positive” results associated with the rush of common disease / common mutation results that were not reproducible (Terwilliger et al., Curr. Opin. Genet. Devel. 12: 726 -734, 2002). Several tests are appropriate to control for the effects of population structure (Cockerham, Evolution 23: 72-83, 1969; Cockerham, 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 divided into two main categories-the genomic control method (Devlin and Roeder, Biometrics 55: 997-1004, 1999) and the structured association (SA) method (Pritchard). And Donelly, Theor. Popul. Biol. 60: 227-237, 2001). Both methods require genotyping a panel of unlinked markers to estimate and correct for the effects of genetic structure, but they are usually applied to sample collections. However, if a pool of samples fails such an assay, it is usually not clear which samples should be removed to correct the problem. An equally plaguing problem with this method is that it is often applied to research samples after the creation of expensive data, and accordingly produces circular logic problems in addition to economic problems; these methods are usually Used to derive information about the population structure using the characteristics of the data for which the association is searched.

  The structure or mixture should not be used as a statistical fuel, and cases and controls are matched and homogeneous in composition to minimize the impact of population structure from the beginning of attempts to identify phenotypically active loci It is generally desirable to qualify samples based on a rough population stratification such as BGA so that For example, ensuring an equal proportion or “race uniformity” within and between cases and controls is not uncommon in the practice of case-control methods. However, for most research purposes, the subjective methods used to measure group affiliation are unsatisfactory. Currently measured using a biogeographic questionnaire, but little knowledge of the population structure other than what is apparent is available, and only the basic relationship between population structure and drug response is obvious. And / or can be controlled. Consistency is an important issue for racial self-reporting in questionnaires and what the Food and Drug Administration intends to submit during the clinical trial design process. However, with such subjective and inaccurate methods of data collection, consistency can be a difficult goal to achieve.

  Rather than reformulating what questions are asked in the questionnaire, consistency can be better addressed by replacing the subjective nature of execution with objective and reproducible scientific methods. Standardization and objectivity are paramount for the collection of racial data, as its measurement can be as subjective as any other human attribute measurement. Racial self-reporting is not as trivial as gender self-reporting, and many people do not know their race or are mixed enough that they bother to classify themselves into a single group It is. Such a scenario is especially common in countries such as the United States where many cultures are combined by immigrants. For example, mostly sub-Saharan African women grew up in Puerto Rico, but may describe themselves as Hispanic. She identifies herself and socio-culturally, but her xenobiotic metabolism and drug target polymorphisms may be related to what is shared among other sub-Saharan people. taller than. By using non-anthropological names that describe the socio-cultural composition of society, current guidelines for taking into account racial information in the research design process can result in poor predictive power and false positive results . Where people grow up, live, and the cultural or sociological habits they adhere to influence how they respond to drugs or the tendency to develop disease there is a possibility. Thus, although non-biological metrics are required, the evidence is that BGA is also influential and therefore needs to be measured in a scientifically accurate and reproducible manner. It suggests.

  Genetic markers present in human DNA provide the best opportunity to reliably measure BGA per individual and have long since been recognized as possible. For example, Reed (Science 244: 575-576, 1973) and Neel (Mutat. Res. 26: 319-328, 1974) call such markers "private" and call them mutation rates. Used to estimate. Reed (supra, 1973) uses the term `` ideal '' (with respect to the use of markers in individual ancestry estimation) to describe hypothetical genetic marker loci in which different alleles are fixed in different populations. Using. Chakraborty et al. (Ethnic. Dis. 1: 245-256, 1991) referred to variants found in only one population as `` unique alleles '' and how the allele frequency was reversed. It has been shown whether a population or BGA affiliation likelihood estimate can be given. “Intrinsic alleles” that are most useful for BGA inference are also those with large differences in allele frequencies between populations (Reed), supra, 1973; Chakraborty et al., Genetics 130: 231-243, 1992; Amer. J. Hum. Genet. 55: 809-824, 1994), and now “ancestral information marker” (AIM); Shriver et al., Hum. Genet. 112: 387-399, 2003; Frudakis et al. , J. Forens. Sci. 48 (4) 771-782, 2003), but "population specific alleles" (PSA, Shriver et al., Amer. J. Hum. Genet. 60: 957-964 1997; Parra et al., Amer. J. Hum. Genet. 63: 1839-1851, 1998)).

  Within the forensic field, statistical methods that use simple serial iterations (STR) to infer the highest level of ancestry (mostly BGA using proportional ancestry notation) in a particular individual, mostly estimate BGA affiliation Can be quite tough to do. STR testing can effectively determine most ancestral origins in most cases, but numbers (5-10%) that do not meet the criteria of classification are ambiguous. Sampling errors caused by rare alleles and STRs were not selected from the genome as their ability to determine population affiliation (i.e., STR allele frequency differentials are not necessarily optimally informative for this purpose as well. Apart from the fact that the main reason for the high level of ambiguity is likely due to mixing, it is clearly a factor of genetic variation for many human populations (Parra et al. 1998, Cavalli-Sforza and Bodmer, The genetics of human populations (see Dover Publications, NY; see pages 387-507) 1999; Rosenberg et al., Supra, 2002). For a given study design, whether to use self-reported information or to use DNA marker testing, and to solve pharmacogenomic or forensic problems, a single patient Categorizing into a group of people sacrifices sensitive but not trivial information about population structure and substructure; for example, 50% African and 50% European affiliations There is no tolerance to allocate. Unfortunately, markers and methods have not yet been described to allow accurate inferences about BGA for more than two groups at a time for individuals. Thus, a need exists for robust markers that are useful for inferring BGA, and for methods of identifying and using such markers. The present invention satisfies this need and provides further advantages.

  The invention is derived, for example, with respect to an individual's ancestry, pigmentation trait, drug responsiveness, and disease susceptibility, as disclosed herein, having a desired predetermined confidence level within an individual population structure. Methods and compositions for measurements that allow inference are provided. By way of example, the methods and compositions were used in a forensic position to examine DNA samples obtained at a serial murder / rape crime scene in Louisiana. Based on psychological profiling, the police had the belief that serial killers were Caucasian men and examined the DNA of over 1,000 Caucasian men, but found no match. As such, the police have relied on the inventors and, using the compositions and methods of the present invention, the offending individuals are African-Americans, and more specifically, 85% sub-Saharan. We decided to have proportional and reliable ancestry of South Africans and 15% Native Americans. Based on this result and the additional results disclosed herein, the average African American has 20% Indo-European ancestry, and a greater level of Indo-European ancestry correlates with a whiter skin tone. And, therefore, the police further advised that the offenders are likely to be African Americans from average to blacker than average skin tone. Within two months of changing efforts based on this information, police arrested African Americans with average skin tone (as African Americans); DNA testing showed that he Determined to be a person found in

  Therefore, the present invention relates to a method for inferring individual traits with a predetermined confidence level. Such methods include, for example, contacting a sample containing a nucleic acid molecule of a test individual with a hybridizing oligonucleotide, wherein the hybridizing nucleotide comprises at least about 10 ancestors exhibiting a population structure correlated with the trait. A single nucleotide polymorphism (SNP) nucleotide occurrence of a panel of informational markers (AIM) can be detected, and its contact stage is suitable for detecting an individual's AIM nucleotide occurrence with a hybridizing oligonucleotide Identifying a population structure that correlates with the nucleotide occurrence of AIM in an individual with a predetermined confidence level, wherein the population structure correlates with a trait. As disclosed herein, at least about 10 AIMs (e.g., 8, 9, 10, 11, 12, 13, 14, 15, 20, 25, 30 or more panels) are examined in carrying out the method of the invention. In general, the greater the number of AIMs examined, the higher the confidence level of inferences made using this method.

  Traits for which reasoning is made by the method of the present invention are related to traits that are known or likely to exist for ethnic predispositions, and whether or not there are ethnic predispositions It can be any trait, including traits that are unknown or unknown. In one embodiment, the trait is biogeographic ancestry (BGA). In one aspect, the panel of AIMs used to examine BGA comprises the AIMs shown in SEQ ID NOs: 1-71. In another aspect, the panel is in SEQ ID NO: 7, 21, 23, 27, 45, 54, 59, 63 and 72-152; SEQ ID NO: 3, 8, 9, 11, 12, 33, 40, 59, 63 and 153 to 239; or SEQ ID NOs: 1, 8, 11, 21, 24, 40, 172 and 240 to 331, in addition to the panels shown in SEQ ID NOs: 1 to 331 Including AIM combinations. As disclosed herein, AIMs useful for carrying out the methods of the invention are linked to a gene associated with a trait (i.e., a gene known to be associated with a trait phenotype). Although not necessary, it is generally not in linkage disequilibrium with the gene (or locus). For example, an AIM useful for inferring an individual's drug responsiveness by the method of the present invention is a gene associated with drug responsiveness (eg, a drug metabolic gene such as a cytochrome P450 gene or a P-glycoprotein gene). Or a drug transport gene). Similarly, an AIM useful for inferring an individual's pigmentation trait by the methods of the invention need not be linked to a gene associated with pigmentation (eg, 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 the panel is associated with the trait that is to be inferred. It is not linked to the gene.

  If BGA is the trait to be inferred, the individuals to be examined are, for example, sub-Saharan African ancestry, Native American ancestry, Indo-European ancestry, East Asian ancestry, Middle East An ancestry can include ancestry, Pacific Islander ancestry, or ancestry that includes any one or combination of ancestry groups, including combinations that include one or more of these ancestry. As such, the proportional ancestry of an individual can include one ancestry (eg, 100% Indo-European ancestry) or any proportion of two, three, four or more ancestry groups. As such, a test individual (or an individual of known proportional ancestry) may include, for example, a sub-Saharan African ancestry and a proportion of two other ancestry groups, or a sub-Saharan African ancestry and an Indo-European ancestry group, and Third ancestry; or Native American and IndoEuropean ancestry and third ancestry; or East Asian and Native American ancestry and third ancestry; or IndoEuropean and East Asian ancestry and third Or a percentage of Native American, East Asian and Indo-European ancestry groups, or sub-Saharan African, Native American and Indo-European ancestry groups.

  In another embodiment, the trait of the test individual that is to be inferred is the individual's responsiveness to drugs, particularly therapeutic drugs. As such, the method of the present invention provides a tool for realizing personalized medicine. Drugs that can be inferred as to whether the test individual responds positively or passively are useful for maintaining or lowering cholesterol levels, for example, cancer chemotherapeutic agents such as paclitaxel. It can be a drug like a possible statin. In one aspect of this embodiment, the AIM of the panel of AIMs used to practice the method comprises an AIM of a gene other than a gene known to be associated with melanin synthesis or metabolism.

  In yet another embodiment, the trait of the test individual that is to be inferred is the individual's susceptibility or predisposition to the disease. As disclosed herein, various traits are associated with population structure at the continental level, while other traits are associated with population structure at a fine level. As such, the methods of the present invention are known to have an ethnic predisposition (i.e., more frequently occur in individuals of a particular ethnic / ancestry group) Diabetes, hypertension And for diseases such as cancer, plus alcoholism that has no ethnic predisposition (or at least is not known to have), or diseases such as schizophrenia, Parkinson's disease and other neurological diseases Means can be provided for inferring on traits such as disease susceptibility.

  In yet another embodiment, the trait of the individual to be inferred is a pigmentation trait. The pigmentation trait can be any such trait including, for example, eye color or darkness, skin color, hair color, or combinations thereof. In one aspect of this embodiment, the AIM of the panel of AIMs used to perform the method is other than a gene known to be associated with melanin synthesis or metabolism, or other aspects of pigmentation. Contains the gene AIM.

  The method of inferring a trait of a test individual by measuring a population structure that correlates with the occurrence of AIM nucleotides in the individual may further comprise identifying a subpopulation structure of the population structure with a predetermined confidence level, wherein the subpopulation structure is Correlates with traits. For example, the population structure of an individual can be correlated by reasoning to an intercontinental group that the individual shares ancestry, such as Indo-European, and the subpopulation structure is a group within the continent where the individual shares the Indo-European ancestry Can be further correlated with, for example, the ethnicity of the Mediterranean.

  Hybridizing oligonucleotides useful in the methods of the invention can be oligonucleotide probes or oligonucleotide primers. Oligonucleotide probes useful in this method can hybridize to a nucleotide sequence comprising a SNP position for AIM, and the nucleotide at the position of the hybridizing oligonucleotide corresponding to the SNP position for AIM is at the SNP position. Either consistent or inconsistent with nucleotide occurrence. Additional oligonucleotide probes useful in the methods of the invention may include nucleotides that hybridize to polynucleotide sequences adjacent and upstream and / or adjacent and downstream to the SNP position and corresponding to the nucleotide position of the SNP. However, it is not necessary and includes oligonucleotide probes that, if such corresponding nucleotides are present in the probe, may be consistent with nucleotide occurrences in the SNP, but are not required.

  Oligonucleotide primers useful in the methods of the present invention include oligonucleotide primers useful for primer extension reactions, as well as oligonucleotide primers that allow amplification of template polynucleotides containing AIM. Such amplification primer pairs generally comprise a forward primer and a reverse primer useful for amplification of a template polynucleotide comprising the AIM of interest. However, two, three, four or more different forward primers can be used for different template polynucleotides containing AIM (e.g., in a multiplex reaction) and common gene sequences (e.g., AIM of a family of related gene sequences). It will be appreciated that it can be used with a common reverse primer for amplification or to generate amplification products of different sizes from a single template. Similarly, one common forward primer can be used with one or more different reverse primers.

  Thus, in one embodiment, the methods of the invention are performed using oligonucleotide primers. In one aspect of this embodiment, the method comprises contacting the sample with an oligonucleotide primer and a polymerase under conditions suitable for the generation of primer extension products. In such a method, the nucleotide occurrence of the SNP is detected by detecting the presence of the primer extension product or by sequencing the primer extension product (or its product) and identifying the nucleotide at the position corresponding to the position of the SNP. Can be measured. In another aspect of this embodiment, the method comprises contacting the sample with an oligonucleotide primer comprising an amplification primer pair and a polymerase under conditions suitable for the generation of an amplification product. In such a method, the nucleotide occurrence of the SNP is detected by detecting the presence of the amplified product or by sequencing the amplified product (or their products) and identifying the nucleotide at the position corresponding to the position of the SNP. Can be measured.

  The method of the present invention is particularly compatible with being performed in a high throughput format, including in a multiplex format, and thus a large number of AIM and / or multiple test individual samples, as well as a control in parallel. Enable inspection. As such, the method may be performed using a format in which the sample to be examined is arranged in an array, particularly an addressable array, for example, on a well of a tray or on a glass slide or silicon chip, It can be partially or fully automated using robotics. If multiple platforms are used, the AIM examined need not necessarily have the largest delta value for a particular trait, for example, with an AIM other than the target AIM for which the hybridizing oligonucleotide is designed. If not substantially cross-hybridized, select AIM so that hybridizing oligonucleotides (eg, amplification primer pairs) can be designed that can be used in a single reaction to examine a panel of AIMs Thus, it appears to be recognized that delta values in multiple sets can also be selected to balance primer suitability.

  The invention also relates to a method for estimating a proportional ancestor of at least two ancestry groups of a test individual with a predetermined confidence level. Such methods include, for example, hybridizing oligos capable of detecting the nucleotide occurrence of a SNP in a panel of at least about 10 AIMs representing a BGA for each ancestral group being examined, including a nucleic acid molecule of a test individual Contacting the nucleotides, wherein the contacting step is under conditions suitable for detecting the nucleotide appearance of the AIM of the test individual by the hybridizing oligonucleotide; and each AIM of the ancestral group being examined Identifying a population structure with a predetermined confidence level that correlates or is most likely to be the nucleotide occurrence of the population structure, wherein the population structure indicates a proportional ancestry. sell.

  Proportional ancestry estimated by the method of the present invention is a percentage of any ancestry group including, for example, sub-Saharan African, Native American, Indo-European, East Asian, Middle Eastern or Pacific Islander ancestry groups. Well, generally, a combination of two or more such ancestry groups. Thus, the proportional ancestry of a test individual may include the proportion of sub-Saharan African and Indo-European ancestry groups (eg, 80% sub-Saharan and 20% Indo-European; or 60% sub-Saharan Africans). , 20% Indo-European, and 20% third ancestry groups); or Native American and Indo-European ancestry groups; East Asian and Indigenous American ancestry groups; Indo-European and East Asian ancestry groups, etc. May include percentage. Similarly, proportional ancestry includes Native American, East Asian and Indo-European ancestry groups; Sub-Saharan African, Native American and Indo-European ancestry groups; Sub-Saharan African, Native American and East Asian ancestry It can include proportions such as groups.

  A panel of AIMs useful for estimating an individual's proportional ancestry is the AIM shown in SEQ ID NOs: 1-331, for example, proportional including Indo-European, Sub-Saharan African, East Asian and Native Americans AIM set forth in SEQ ID NOs: 1-71 which may be useful for measuring ancestry; or SEQ ID NOs: 7, 21, which may be useful for measuring proportional ancestry of East Asians and sub-Saharan Africans 23, 27, 45, 54, 59, 63 and 72-152; or SEQ ID NOs: 3, 8, 9, 11, 12 which may be useful for measuring the proportional ancestry of East Asians and Indo-Europeans 33, 40, 59, 63 and 153-239; or SEQ ID NOs: 1, 8, 11, 21, 24, 40, which may be useful for determining the proportional ancestry of Indo-Europeans and Sub-Saharan Africans 172 and 240-331, AIM.

  In one embodiment, an estimate is made where the proportional ancestry includes a proportion of three ancestry groups. In one aspect of this embodiment, identifying a population structure that is most likely correlated with or likely to be a trend in the AIM nucleotide appearance of the tested individual is a sub-Saharan African ancestry group, indigenous American ancestry Determining the likelihood of affiliation for each of the group, Indo-European ancestry group, and East Asian ancestry group; then selecting the three ancestry groups with the largest likelihood values; Determining the likelihood of all possible proportional affiliations among the three ancestral groups having, thereby identifying a population structure or proportional affiliation that correlates with the nucleotide occurrence of the AIM of the test individual, As well as identifying only one proportional combination of maximum likelihood.

  In another aspect of this embodiment, identifying the population structure that is most likely to correlate with or likely to be the nucleotide occurrence of AIM is Performing six binary comparisons, including likelihood determination for affiliation; then selecting the three ancestry groups with the largest likelihood values; among the three ancestry groups with the largest likelihood values Determining the likelihood of all possible proportional affiliations, whereby the population structure or proportional affiliation most likely correlates with or is likely to be a trend in the AIM nucleotide appearance of the tested individual Identified steps; and identifying only one proportional combination of maximum likelihood.

  In yet another aspect of the embodiment where the proportional ancestry includes three ancestry groups and an estimate is made, the method performs three ternary comparisons between the groups; the largest likelihood value; Determining the likelihood of all possible proportional affiliations among the three ancestral groups with, thereby correlating with or likely to be correlated with the AIM nucleotide appearance of the test individual Performed by identifying the highest, population structure or proportional affiliation, steps; and identifying only one proportional combination of maximum likelihood. In another aspect of this embodiment, the method may further comprise creating a graphical representation of the comparison of the three ancestry groups, wherein each graphical representation is represented independently by a triangular vertex, The maximum likelihood value of a proportional affiliation for an individual that includes a triangle includes a point within the triangle. If desired, the graphical representation may further include confidence contours that indicate the confidence level associated with estimating proportional ancestry.

  In another embodiment, an estimate is made where the proportional ancestry includes a proportion of four ancestry groups. In various aspects of this embodiment, identifying a population structure that correlates or is most likely to be the trend of AIM nucleotide occurrences in a test individual is a six-way comparison between groups. Perform three-way comparisons, or perform three-way comparisons; determine the likelihood of all possible proportional affiliations in the four ancestor groups with the highest likelihood values A population structure or proportional affiliation that is most likely to correlate with or is likely to be a trend in the AIM nucleotide appearance of the test individual; and only the maximum likelihood This is done by identifying one proportional combination. In one aspect of this embodiment, the method may further comprise creating a graphical representation of the comparison of the three ancestry groups, wherein the graphical representation is a pyramid, wherein each ancestor group is independently represented by a pyramid vertex. And the maximum likelihood value of proportional affiliation for an individual includes a point in the pyramid. If desired, the graphical representation may further include a confidence contour that includes a sphere centered at that point, where the sphere indicates the confidence level associated with estimating the proportional ancestry.

  Estimating the proportional ancestry of at least two ancestor groups of a test individual with a certain confidence level by identifying a population structure that exhibits proportional ancestry is associated with one of the ancestry groups for which the test individual has a proportional ancestry. The method may further include the step of identifying a sub-group structure exhibiting a certain ethnicity. This method identifies a subpopulation structure of the population structure that correlates with the occurrence of AIM nucleotides in the test individual, and the subpopulation structure correlates with the ethnicity of the test individual. Such methods of identifying subpopulation structures include, for example, an AIM that indicates affiliation for a biogeographic ancestry group (individuals are proportionally associated with more than one biogeographic ancestry group). Identifying those chromosomes of the test individual comprising, contacting a sample containing the nucleic acid molecule of the test individual with a second hybridizing oligonucleotide capable of detecting the nucleotide appearance of the SNP in the second panel of AIM Stage, AIM in the second panel contains information about the ethnicity within one of these groups and includes an AIM that indicates the larger (intercontinental) ancestral group in which that ethnicity occurs Identifying the subpopulation structure that is present on the same chromosome of the test individual; and correlating with the second panel of AIM nucleotide occurrences, wherein the subpopulation indicates the ethnicity of the test individual's ancestry group Is in the stage Ri may be implemented.

  By such a method, 60% of the test individuals were tested with hybridizing oligonucleotides specific for the first panel of AIM (eg 71 AIMs of 71 exemplified AIMs; SEQ ID NOs: 1-71). Can be determined to be Indo-European (IE) and 40% East Asian. In such cases, only the fraction of all possible AIMs that could represent the IE ancestry group would have been positive (if all were positive, the individual would have been 100% IE) and therefore the individual Only a portion of the chromosome or chromosomal region appears to be of Indo-European origin. The individual's chromosomes containing positive AIM for IE are then identified, and a second hybridizing oligonucleotide specific for the second panel of AIMs is selected (e.g., 1000 covering all 23 pairs of human chromosomes). (From as many as AIM groups), AIM in the second panel varies highly in allelic frequency among IE ethnic groups, and therefore exhibits IE ethnicity, and also the first panel AIM is IE positive Limited to those present on the chromosomes that were present. A subpopulation structure that correlates with the nucleotide appearance of AIM in the second panel is then identified, and accordingly, ethnicity with respect to the IE ancestry group of the test individual, eg, the IE ancestry group is Northern Europe, Mediterranean, It is derived from the ethnicity of the Middle East or South Asia India. As such, the method includes an AIM that correlates with a population structure that represents IndoEuropean biogeographical ancestry, and that further includes an AIM that more specifically correlates with a subpopulation structure that represents the ethnicity of the Mediterranean race. Provides a means for identifying the chromosomal ethnic origin (eg, the Mediterranean origin of a chromosome previously determined to be of Indo-European origin).

  In another embodiment, a method for estimating a proportional ancestry of a test individual can include creating a global ancestor map, wherein the location of a population having a proportional ancestry corresponding to the proportional ancestry of the test individual is an ancestor map. Shown above. As such, the method can supplement the system information. For example, the method involves overlaying an ancestor map with a phylogenetic map, where the pedigree map indicates the location of a population that has geopolitical relevance with respect to the examined individual, and the most likely family of the examined individual The method may further include statistically combining the information of the ancestor map and the phylogenetic map so that an estimation is obtained.

  By the method of the present invention, identifying a population structure that is most likely to correlate with or likely to be a trend in AIM nucleotide occurrences, the tester's AIM nucleotide appearance is indicative of BGA AIM nucleotide occurrences. By comparing to a known proportional ancestor corresponding to. Known proportional ancestry corresponding to the nucleotide occurrence of AIM indicating BGA may be included in the table or other list, and the nucleotide occurrence of the test individual may be visually compared to the table or list, or included in the database In other words, the comparison can be made electronically, for example, using a computer. In addition, each known proportional ancestor corresponding to the occurrence of AIM nucleotides representing BGA is associated with a photograph of the person for whom the known proportional ancestor was determined, and further infers the physical characteristics of the test individual accordingly Means can be provided. In one aspect, the photograph is a digital photograph and includes digital information that may be included in a database that may further include a plurality of such digital information of the digital photograph, each of which represents an AIM nucleotide occurrence that indicates a person's BGA Associated with a known proportional ancestor.

  In another aspect, the method of the present invention may further comprise identifying a photograph of a person having a proportional ancestry that corresponds to the proportional ancestry of the test individual. Such identification can be done by manually examining one or more files of the photograph, which are organized according to, for example, the nucleotide occurrence of a person's AIM in the photograph. Identifying the photos also includes scanning a database containing multiple files, each file containing digital information corresponding to a digital photo of a person with a known proportional ancestor, and the BGA of the individual being examined. This may be done by identifying at least one photograph of a person with an AIM nucleotide occurrence showing a BGA that matches the indicated AIM nucleotide appearance.

  Thus, the present invention also provides a product that is at least one photograph of one person with a known proportional ancestry that corresponds to a population structure that includes nucleotide occurrences of AIM representing BGA, and each product of the plurality Relates to a plurality of such articles containing one (or more) photograph of a person with a known proportional ancestry corresponding to a population structure containing nucleotide occurrences of AIM indicating BGA. Goods can be contained in one file, or multiple goods can be contained in one file, for example, one file contains multiple photos of different people, some or all of those people have a BGA With the same or different known proportional ancestry corresponding to the population structure containing the nucleotide occurrences of the indicated AIM.

  Thus, multiple such items are provided and multiple files are also provided, each file having the same or different known proportional ancestry corresponding to the population structure containing the AIM nucleotide occurrences representing BGA. It may include one or more items, i.e. photographs, that may be of one or more people. For example, the plurality of different files may each contain one (or more) photograph of a person with a known proportional ancestry that corresponds to a population structure containing AIM nucleotide occurrences representing BGA. The plurality of different files may also include photographs of two or more different people having the same or substantially the same proportional ancestry, each corresponding to a population structure containing AIM nucleotide occurrences representing BGA. As such, if each file contains one or more photos of one or more people and one or more photos of two or more different people, the different people are the same Or it may contain files that may have different known proportional ancestry.

  In one embodiment, the product, ie a photograph of a person with a known proportional ancestry corresponding to a population structure that includes nucleotide occurrences of AIM indicative of BGA, is a digital photograph containing digital information. As such, the digital information of the present invention, or the digital information of multiple digital photo products, may be included in the database. As such, the present invention further provides a plurality of products, each including at least two digital photographs containing digital information. In one aspect of this embodiment, digital information about one or more of the items is contained in a database, which is, for example, computer hardware or software, magnetic tape, or a computer such as a floppy disk, CD or DVD. It can be included on any medium suitable for containing such a database, including disks. As such, the database can be contained by the computer, can contain the database, can accept the medium containing the database, or can access the database through a wired or wireless network, such as an intranet or the Internet. Can be accessed.

  The present invention also includes a plurality of hybridizing oligonucleotides, each hybridizing oligonucleotide including at least 15 consecutive nucleotides of the polynucleotide shown in SEQ ID NOs: 1-331, and the plurality Relates to a kit comprising at least 5 of such oligonucleotides, each based on a different polynucleotide set forth in SEQ ID NOs: 1-331. In one embodiment, the hybridizing oligonucleotide comprises at least 15 of at least 5 of the polynucleotide set forth in SEQ ID NOs: 1-71 or the polynucleotide complementary to any of SEQ ID NOs: 1-71. Contains consecutive nucleotides.

  The hybridizing oligonucleotides of the kits of the invention can include probes that are useful for detecting specific AIMs, including specific nucleotide occurrences at ANP SNP or DIP (deletion / insertion polymorphism) positions. May include primers, including primers useful for primer extension reactions and primer pairs useful for nucleic acid amplification reactions; or may include combinations of such probes and primers. The hybridizing oligonucleotide of the plurality is a nucleotide position of AIM (eg, nucleotide position 50 of any of SEQ ID NOs: 1-34 and most others, nucleotide position 56 of SEQ ID NO: 35, SEQ ID NO: 50 nucleotides 44, or nucleotides 26 of SEQ ID NO: 56) or nucleotides corresponding to their complementary nucleotide sequence, such hybridizing oligonucleotides appear at a particular nucleotide at the SNP position of AIM It is useful as a probe for identifying the presence or absence of.

  In another embodiment, the kit comprises at least one pair of hybridizing oligonucleotides useful for detecting nucleotide occurrence at the SNP (or DIP) position of AIM. In one aspect of this embodiment, the pair of hybridizing oligonucleotides is adjacent to and downstream of the SNP (or DIP) position of one oligonucleotide and AIM that hybridizes upstream and downstream of the SNP position of AIM. Includes a second oligonucleotide that hybridizes, and one or the other of the pair is complementary to the nucleotide occurrence suspected of being in the SNP (or DIP) position of AIM (i.e., one of the polymorphic nucleotides) Such a pair of hybridizing oligonucleotides further comprising nucleotides are useful in oligonucleotide ligation assays. In another aspect of this embodiment, the pair of hybridizing oligonucleotides comprises an amplification primer pair comprising a forward primer and a reverse primer, and such a pair of hybridizing oligonucleotides comprises an AIM SNP (or Useful for amplifying polynucleotide portions containing DIP) positions.

  The kit of the present invention may further comprise additional reagents useful for performing the method of the present invention. As such, the kit includes, for example, one or more polynucleotides comprising AIM, including a polynucleotide comprising AIM that is designed to be detected by a hybridizing oligonucleotide or a pair of hybridizing oligonucleotides of the kit. And such polynucleotides are useful as controls. Further, the hybridizing oligonucleotides of the kit can be detectably labeled, or the kit comprises one of the kits or different detectable labels that can be used to label the hybridizing oligonucleotide differently. Reagents useful for detectably labeling a plurality of hybridizing oligonucleotides can be included; such kits are for linking a label to a hybridizing oligonucleotide or for detecting a labeled oligonucleotide. And the like. The kits of the invention also include ligases, for example, when the hybridizing oligonucleotide of the kit includes a primer or amplification primer pair, a polymerase; or when the kit includes a hybridizing oligonucleotide useful for oligonucleotide ligation assays. , May be included. In addition, the kit can include appropriate buffers, deoxyribonucleotide triphosphates, and the like, depending, for example, on the specific hybridizing oligonucleotides included in the kit and the purpose for which the kit is to be supplied.

In a mixed population, a diagram is provided that shows the manner in which chromosomal segments are recombined over time. Initially, the parent population has chromosomal segments that are continuous with respect to AIM along the segment. In the first hybrid (F1), every person has one complete chromosomal segment from each parental population. Much more combinations are possible in the F2 generation. The relative likelihood of the non-recombinant versus recombinant genotype shown in F2 depends on the size of the chromosomal segment. A segment approximately the size of a human chromosome appears to produce an average of several recombination events in a single meiosis (a single recombination is 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) xF1 illustrates the genotype of a person with one F (N) parent and one F1 parent; and F (N) xF2 represents one F (N) parent and one person Illustrates the genotypes of people with F2 parents. 3 shows a triangular graph created using the algorithm described in Example 6 (see also Table 12). NAM, Native American; AFR, Sub-Saharan African; EUR, Indo-European. FIG. 2A illustrates the extension of the line from the NAM vertex of the triangle to the opposite side, with the opposite side representing 0% Native American ancestry. A circle is shown at the position of the estimated proportional ancestry (see FIG. 2B), and the hatch on the line indicates the percentage of indigenous American ancestry (approximately 15%). FIG. 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 percentage of each respective ancestor; ie 15% Native Americans, 60% Indo-Europeans, and 25% Africans. Figure 3 shows a triangular plot depicting one approach illustrating the value and accuracy of an individual's ancestry estimates. The typical distribution of three populations is shown (European Americans: black squares; African Americans: open triangles; and African / Indigenous American populations: open circles). A single individual is also shown, with confidence intervals represented as concentric rings that enclose point estimates (solid circles). Like the topological map, each concentric ring represents a decrease in likelihood in units of 1 log unit (probably less than 10 times). In this example, the individual has a likelihood interval space that is symmetric and circular. The interval space takes many forms depending on the mixing ratio of the questioned objects and the allele frequency of the typed marker. Three African American samples (open circles: WASH-Washington DC, AFCAR-African Caribbean and BOG-Bogarousa), European American samples (open circle: SCO-State College), and Spanish American samples Provide a triangle plot showing the mean mixture estimate for (open diamonds: SLV-San Luis Valley CO). In parentheses, the genetic contributions to the average African (AFR), Indo-European (EUR) and Native American (NAM) samples are shown. Shows the genetic structure in the US resident population. FIG. 5A shows the percentage of unlinked AIM showing a significant association. Expected values are based on the 5% significance level. Values for the Washington DC sample are based on 33 AIMs, 19 AIMs for San Luis Valley CO, and 34 AIMs for State College PA. FIG. 5B shows the correlation between individual ancestry estimates based on independent subsets of informative markers. The average correlation is based on 100 replicates. The total number of markers is the same as for FIG. 5A. The corresponding p-value is shown at the bottom of the graph. A triangular plot for the father (Figure 6A) and mother (Figure 6B) is shown. FIG. 6 shows a triangular plot for each of the three father and mother children represented in FIG. The distribution of AIM in the genome is shown (Chrom. Number, chromosome number). Demonstrate the toughness of BGA mixing ratio analysis using AIM (see Example 2). The confidence (contour) of the maximum likelihood estimate (MLE) is affected as expected by the removal of AIMs that inform the specific pairwise comparison. The first contour extending from the MLE defines a triangular plot space with a likelihood two times lower than that of the MLE, and the second contour defines a space with a likelihood five times lower than the MLE. FIG. 9A shows the MLE and confidence contours obtained with 71 AIMs; the actual percentages are shown. FIG. 9B shows the results obtained after removing those AIMs that were used to obtain the results shown in FIG. 9A from an informed analysis of East Asian-Indigenous American distinctions. MLE is relatively unaffected and the confidence contours along the East Asian-Indo-European (European) axis and the Native American-European axis remain undistorted, but the confidence contours are Distorted along the East Asian-Indigenous American axis. The BGA mixing ratio measured for 8 individuals in the family tree is shown. Circles represent women, squares represent men, and BGA affiliation for each individual is shown as a fraction where the numerator represents IndoEuropean BGA and the denominator represents Native American BGA. Except for what is indicated by an asterisk (*) indicating that the individual was measured to have 4% East Asian BGA, all sub-Saharan African BGAs and East Asian BGAs Was not included. All others show a family tree demonstrating how Chinese great-grandfathers in an Indo-European family tree can produce grandchildren with Indo-European / East Asian ancestry. Individuals who are 100% East Asian (Chinese) are shaded; mixed results for the bottom male (square) (short arrow) in the family tree. The grandmother indicated by the long arrow is about 50% / 50% East Asian / Indo-European mixed, and her daughter, the subject's mother, is 25% / 75% East Asian / Indian European mixed. Expected to be (Example 3). Figure 5 shows the distribution of all SNPs effective for genotyping by chromosome arm for a group of patients treated for elevated cholesterol levels. Caucasian individuals taking Lipitor ™ with known responses for cholesterol (lip TC), low density lipoprotein (lip LDL), liver transaminase AST-SGOT (lip SGOT) and ALT-GPT (lip GPT) measurements The distribution of SNP at (n = 180) is shown. SNPs with significant (> 0.20) delta values in various trait classes were selected. For example, in about 70% of patients, Lipitor ™ caused a decrease in LDL. For any given SNP, the delta value (δ) is the difference in the minor allele frequency among individuals whose LDL has been reduced by at least 20% versus those whose LDL has not changed. Same as for FIG. 13 except that response was measured after treatment with Zocor ™ (n = 150) and only total cholesterol (zoc TC) and LDL (zoc LDL) were examined The analysis of is shown. The distribution of SNPs in the chromosome (δ> 0.11) for 1,000 individuals of known eye color is shown. Shown is the distribution of SNPs in the chromosome (δ> 0.11) for 1,000 individuals of known hair and eye color.

DETAILED DESCRIPTION OF THE INVENTION The present invention is based on the identification of ancestral informative markers (AIM) that are useful to infer the level of population structure of an individual, which in turn allows inferences about the various traits of the individual. Furthermore, the AIM of the present invention has been demonstrated that markers correlate with traits, whether or not they are in linkage disequilibrium with genes or loci known to be associated with traits. Yes. As such, when linked to a trait, i.e., the marker has a low crossover percentage, for example with respect to a gene (or locus) known to be associated with (or associated with) the trait The previously described markers considered useful only when in physical proximity to a gene known to be associated with a trait such that it is particularly characterized, and the AIM of the present invention Can be distinguished. In contrast, there is no requirement that markers (AIMs) useful in this method be in linkage disequilibrium with the gene / trait, and in fact, the AIMs disclosed herein as being correlated with traits , And a gene / locus known to be associated with a trait may be located on a different chromosome.

  AIM is a locus that represents an allele with a high frequency difference between populations. AIM is generally exemplified herein by single nucleotide polymorphisms (SNP; see, eg, SEQ ID NO: 1), and deletion / insertion polymorphisms (DIP; see, eg, SEQ ID NO: 363). ing. As disclosed herein, AIM is practiced at the population level (with respect to race), at the subgroup level (with respect to ethnicity), and at the microgroup level (with respect to families within ethnic groups). Can be used to estimate the biogeographic ancestry (BGA) of an individual or a collection of individuals at a typical, phenotypically qualified level (eg, cases and controls). Such ancestry estimates at the subgroup and individual levels, for example, include genetic phenotypes that differ qualitatively or in frequency between populations, including the likelihood that an individual will respond to a particular drug or the propensity of an individual to develop a disease Can be directly instructive about the nature. Ancestral estimation can also provide an essential basis for the use of mixed mapping (AM) methods to identify the genes underlying these traits.

As illustrated herein, a panel of 71 AIMs (SEQ ID NOs: 1-71) was identified from a survey of over 800 candidate AIMs (see also SEQ ID NOs: 72-331), A method was developed to investigate these AIMs as a means to obtain an accurate estimate of proportional ancestry. The methods and markers of the present invention have been validated in studies using skin pigmentation as a model phenotype (see also WO 02/097047 (PCT / US02 / 16789), which is hereby incorporated by reference. Incorporated in). The first markers were genotyped in two population samples, mainly including African ancestry, in African American samples from Washington DC and African Caribbean samples from England, and in European American samples from Pennsylvania. (See Example 1). In two African population samples, a very strong correlation was observed between estimation of individual ancestry and skin pigmentation as measured by reflectance measurements (R ² = 0.21 for African American samples) , P <0.0001, and R ² = 0.16, p <0.0001 for a sample of British African Caribbean. These correlations confirm the validity of the ancestor estimates and also identify high-level population structures associated with mixing that can be detected using other tests that characterize these populations and identify their genetic structure showed that. These results demonstrate that estimation of an individual's ancestry can be made based on DNA analysis using a relatively small number of well-defined genetic markers (AIM).

  The methods and genetic markers disclosed herein include, for example, 1) for estimation of ancestry proportions in individuals from their DNA; 2) genetic structure as a control for research designs commonly used for genetic investigations. 3) To build a physical profile by inferring ancestor-related features that can be meaningful in forensic research; 4) Disease predisposition called “mapping by ancestral linkage disequilibrium” (MALD) And 5) provide tools for several distinct purposes, including to predict a significant portion of individual patient response to prescription drugs and drug sales. As such, the present invention is useful, for example, for 1) examples of statistical methods and uses for measuring ancestry percentages from gene sequences within individuals; 2) measuring ancestry percentages within individuals or research groups As a statistical method, several hundred AIMs selected and identified from publicly available single nucleotide polymorphism (SNP) databases; 3) useful for measuring the proportion of ancestors within an individual or study group Hundreds of AIMs that have been demonstrated to be; and 4) provide software programs that can be used to determine the proportion of ancestors within an individual or study group.

  Efforts to control two sources of population structure, including sampling effects and natural human demography, previously thought to disrupt attempts to identify genetic markers associated with specific traits It has been made. However, as disclosed herein, population structure reflects human demographics, and markers that correlate with trait values are useful as reporters of structures that correlate with trait values (phenotypically active genes (Rather than a marker in the LD for the locus), therefore, provides a valuable tool that allows accurate classification in a cost-effective and practical manner. Alleles associated with traits by population structure are not linked to phenotypically active loci and are simply correlated with trait values because they are enriched in a branch of the human genealogy where trait values are more common. It ’s just that. As disclosed herein, the distribution of trait values among various sects of the human genealogy is such that accurate classification is a correct recognition of its structure rather than a complete understanding of the biological mechanism of the trait. As a result, when considered for use to identify a phenotypically active locus, a marker that was considered a false positive actually did an accurate classification analysis. That is, they are truly positive, provided that the structure from which they are derived is a reflection of human demographics rather than sampling effects. The method is based on the correlation between the marker and BGA, which itself is at a considerable level of complexity correlated with the trait value and is not linked or linkage disequilibrium.

  Therefore, the present invention provides a method for inferring an individual trait with a predetermined confidence level. In one embodiment, the method of the invention comprises subjecting a nucleic acid molecule sample of a test individual to a hybridizing oligonucleotide capable of detecting the nucleotide occurrence of a single nucleotide polymorphism (SNP) of at least about 10 AIM panels. Identifying a population structure that correlates or is most likely to be a trend of AIM nucleotides in an individual, wherein the population structure correlates with a trait Is done in stages. The AIM panel shows population structures that are selected and based on the specific platform used to perform the method, in their delta values (see below) and where relevant. AIM is exemplified herein by the polynucleotides shown as SEQ ID NOs: 1-331, and the SNP position is generally at nucleotide position 50 (but, for example, SEQ ID NO: 35, nucleotide position 56; sequence No: 51, 48; see SEQ ID NO: 56, 26).

  The test individual whose trait is to be inferred can be any individual whose trait is desired to be inferred, and is generally a human. However, the method of the present invention also infers traits of other mammals including, for example, livestock such as cats, dogs or horses; agricultural livestock such as cattle, sheep, pigs or goats; or other animals Can be used to The trait that can be examined can be any trait of interest, including proportional ancestry (BGA); hair, skin, or iris pigmentation; or drug responsiveness, as illustrated herein.

  The method of the present invention is particularly useful because it allows inferences about the desired trait with a predetermined confidence level. As used herein, reference to a “predetermined confidence level” is made using a statistical method in which the inference or estimation of the present invention provides a confidence interval determined for an average or maximum likelihood value. Means. In addition to determining the maximum likelihood value within an individual or sample structure, other similarly likely values can also be determined, which are x-times likelihood confidence intervals (x is 2, 5, or 10 Can be combined to define any number such as For example, all structural results corresponding to likelihood values 10 times lower than the maximum likelihood value can be plotted or listed to define a 10-times likelihood confidence interval. As far as any statistical test is concerned, the assay method of the present invention is designed such that execution of the test results in a value with the desired confidence level. As disclosed herein, the methods of the invention can be performed such that the results have a predetermined confidence level by changing the number of AIMs examined for a trait. For example, the use of a particular panel of 10 AIMs allows an inference to be made regarding whether an individual has a particular trait, such as responsiveness to Lipitor ™, with a particular confidence level However, the use of 20 AIM panels that can be duplicated, though not necessary, in part with 10 AIM panels is the same reasoning but is made with a higher confidence level. enable. Similarly, the use of two panels of 10 AIMs is inferred that an individual has, for example, 80% Indo-European ancestry and 20% East Asian ancestry (with an error, eg, ± 10%) Although the use of two panels of 20 AIMs each is the same reasoning, it can be possible with an error of, for example, ± 5%.

  Samples useful for practicing the methods of the invention include any biological sample of a test individual that contains a nucleic acid molecule, including a portion of a gene sequence that contains the AIM to be examined, or a polymorphism of AIM Can be any biological sample containing the encoded polypeptide that results in an amino acid change in the encoded polypeptide. 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, spinal fluid, and the like.

  Nucleic acid samples useful for practicing the methods of the invention will depend in part on whether the SNP to be identified is in the coding region or in the non-coding region. Where one or more SNPs are present in the non-coding region of a gene, the nucleic acid sample is generally a deoxyribonucleic acid (DNA) sample, particularly genomic DNA or an amplification product thereof. However, where AIM is contained within a transcribed sequence, for example, rDNA, microsatellite DNA, or heteronuclear ribonucleic acid (RNA) containing an unspliced mRNA precursor RNA molecule containing a non-coding RNA sequence. In this case, RNA samples can be examined directly or cDNA or amplification products thereof can be examined by this method. Where one or more SNPs are present in the coding region of a gene, the nucleic acid sample can be DNA or RNA, or a product derived therefrom, eg, an amplification product. Furthermore, although the method of the invention is illustrated with respect to nucleic acid samples, if a particular SNP is present in the coding region of a gene, the result is a position at a position corresponding to the SNP due to non-degenerate codon changes. It will be appreciated that polypeptides containing different amino acids can be generated. As such, in one aspect, the methods of the invention are performed using a sample containing the polypeptide of interest.

  The method of the invention is performed by contacting the sample and hybridizing the oligonucleotide under conditions suitable for detecting the nucleotide appearance of the individual's AIM by the hybridizing oligonucleotide. Further, in an aspect of the method of the invention, the sample can be contacted with a second hybridizing oligonucleotide, for example, to determine the substructure. The term "second" is used for convenience of discussion when used with respect to hybridizing oligonucleotides (or a panel of AIMs), for example to allow a clear distinction of the steps for performing the method. That should be recognized. In this regard, it should further be appreciated that one or more hybridizing oligonucleotides used, for example, to determine population structure can also be included in the second hybridizing oligonucleotide. is there.

  Suitable conditions for detecting the nucleotide occurrence of AIM include specific assays that are to be used in addition to the hybridizing oligonucleotide sequence, including length and complementarity, and, for example, assay methods. Depending on whether or not is to be performed as a multiplex assay. Hybridizing oligonucleotides that are at least 15 nucleotides long may include deoxyribonucleotides or ribonucleotides linked together by phosphodiester bonds, which are commonly used in single-stranded form, but are single-stranded or double-stranded. It can be a chain. Such hybridizing oligonucleotides can be prepared using chemical synthesis methods or by enzymatic methods such as polymerase chain reaction (PCR).

  Hybridizing oligonucleotides or other polynucleotides useful in the methods or included in kits of the invention may also contain nucleosides or nucleotide analogs, may have backbone linkages other than phosphodiester linkages, and the like Oligonucleotides offer certain advantages such as increased stability or more desirable hybridization properties. Nucleotide analogs are well known in the art and are polynucleotides containing such nucleotide analogs, but are commercially available (Lin et al., Nucl. Acids Res. 22: 5220-5234, 1994; Jellinek et al., Biochemistry 34: 11363-11372, 1995; Pagratis et al., Nature Biotechnol. 15: 68-73, 1997, each incorporated herein by reference). Covalent bonds also include a number of other bonds, including thiodiester bonds, phosphorothioate bonds, peptide-like bonds, or any other bond known to those skilled in the art as useful for linking nucleotides to make synthetic oligonucleotides. (Eg, Tam et al., Nucl. Acids Res. 22: 977-986, 1994; Ecker and Crooke, BioTechnology 13: 351-360, 1995, each of which is incorporated herein by reference. ) Non-naturally occurring nucleotide analogs or linkages that link nucleotides or analogs include, for example, samples containing tissue culture media or cell extracts, as modified oligonucleotides may be less susceptible to degradation. It can be particularly useful where the oligonucleotide is to be exposed to an environment that may contain nucleolytic activity.

  In general, a hybridizing oligonucleotide useful for the purposes of the present invention is at least about 15 bases long, which is sufficient to allow the oligonucleotide to hybridize selectively to a target polynucleotide comprising AIM. And can be at least about 18 nucleotides in length, or 21 nucleotides in length or 25 nucleotides in length or longer. The term “selective hybridization” or “selectively hybridize” means a moderately stringent or highly stringent physiology that can distinguish related nucleotide sequences from unrelated nucleotide sequences. Refers to hybridization under dynamic conditions. Conditions used to achieve a particular level of stringency in a nucleic acid hybridization reaction include, for example, length, degree of complementation, nucleotide sequence components (eg, relative GC: AT content), and It is known that the type of nucleic acid, ie the oligonucleotide or target nucleic acid sequence, will vary depending on the nature of the nucleic acid to be hybridized, including DNA or RNA. An additional consideration is whether one of the nucleic acids is immobilized on, for example, a filter, bead, chip or other solid matrix.

  Methods for selecting appropriate stringency conditions can be determined empirically or can be estimated using various equations and are well known in the art (see, for example, Sambrook et al., Supra, 1989. reference). Examples of increasingly stringent conditions are: 2X SSC / 0.1% SDS (hybridization conditions) at about room temperature; 0.1% SDS (low stringency conditions) at about room temperature; 0.2X SSC / 0.1% SDS (medium stringency conditions); and 0.1X SSC (high stringency conditions) at about 68 ° C. Washing can be performed using only one of these conditions, e.g., high stringency conditions, or any of the listed steps, e.g., in the order listed above for 10-15 minutes each, or Everything can be repeated and each condition can be used. As such, the final conditions will vary depending on the particular hybridization reaction involved and can be determined empirically. It should be appreciated that a variety of conditions can be utilized to provide selective hybridization conditions. For example, if a multiplex assay is to be performed with multiple different hybridizing oligonucleotides specific for different AIMs in the panel, the conditions (and AIM / hybridizing oligonucleotides) are selected. Selective hybridization can be selected to occur for all hybridizing oligonucleotides in the reaction.

  In various embodiments, it may be useful to detectably label a polynucleotide or hybridizing oligonucleotide. Detectable labeling of polynucleotides is well known in the art and includes, for example, chemiluminescent labels, radionuclides, enzymes, haptens such as digoxigenin and biotin, fluorophores, and unique oligonucleotide sequences. Use of such detectable labels. For example, a PCR product can be performed, but one primer is biotinylated and the other primer contains digoxigenin. The amplification product can then be bound to a streptavidin plate, washed, reacted with an enzyme-linked antibody to digoxigenin, and developed with a chromogenic, fluorogenic or chemiluminescent substrate for the enzyme. Alternatively, a radioactive method may be used to detect the generated amplification product, for example, by including radiolabeled deoxynucleoside triphosphates in the amplification reaction and then blotting the amplification products onto DEAE paper for detection. Can be used. Furthermore, if one primer is biotinylated, streptavidin-coated scintillation proximity assay plates can be used to measure PCR products. Additional detection methods include chemiluminescent labels, e.g., lanthanide chelates such as those used in DELFIA® (Pall Corp.), fluorescent labels, or electrochemiluminescents such as ruthenium trisbipyridyl (ORI-GEN). A label may be used.

  A method for detecting the occurrence of a nucleotide at the SNP or DIP position of AIM selectively comprises one or more oligonucleotide probes or primers that hybridize selectively to a target polynucleotide spanning AIM, for example, including an amplification primer pair. Can be used. Oligonucleotide probes useful in practicing the methods of the invention include, for example, oligonucleotides that are complementary to and span the portion of the target polynucleotide that includes the position of the SNP (or DIP) and are identified at the position of the SNP. Is detected by the presence or absence of selective hybridization of the probe. Such a method involves contacting the target polynucleotide and hybridized oligonucleotide with an endonuclease, and cleaving the probe, depending on whether the nucleotide occurrence at the SNP site is complementary to the corresponding nucleotide of the probe. It may further comprise detecting the presence or absence of the product. A pair of probes that specifically hybridize close to, upstream, and close to and downstream of the SNP site, one of the probes comprising nucleotides complementary to the nucleotide occurrence of the SNP Probes can also be used in oligonucleotide ligation assays, where the presence or absence of a ligation product indicates nucleotide appearance at the SNP site. Oligonucleotides can also be useful as primers, eg, for primer extension reactions, where the product (or absence of product) of the extension reaction indicates nucleotide appearance. In addition, primer pairs useful for amplifying portions of the target polynucleotide containing SNP or DIP sites may be useful, and amplification products may be inserted or deleted to measure nucleotide appearance at the SNP site or at the DIP site. Checked to determine if there is a loss.

  A number of methods are known for measuring nucleotide appearance at a particular position in a polynucleotide (ie, SNP or DIP). Such methods utilize, for example, one or more oligonucleotide probes or primers that selectively hybridize to a target polynucleotide that includes one or more SNP positions, for example, including an amplification primer pair. Can do. Hybridizing oligonucleotides useful in practicing the methods of the invention are complementary to the portion of the target polynucleotide that includes, for example, the position of SNP or DIP (including whether DIP has a deletion or insertion). 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 involves contacting the target polynucleotide and hybridized oligonucleotide with an endonuclease, and cleaving the probe, depending on whether the nucleotide occurrence at the SNP site is complementary to the corresponding nucleotide of the probe. It may further comprise detecting the presence or absence of the product.

  Oligonucleotide ligation assays can also be used to identify nucleotide occurrences at SNP sites, where a pair of probes selectively hybridize close and upstream, and close and downstream to the SNP site, And one of the probes contains a terminal nucleotide complementary to the 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 the upstream and downstream oligonucleotides are ligated in the presence of ligase. As such, the presence or absence of a ligation product indicates the occurrence of a nucleotide at the SNP site.

  Hybridizing oligonucleotides can also be useful as primers, such as for primer extension reactions, where the product of the extension reaction (or absence of product) is a nucleotide occurrence at the SNP site or an insertion or deletion at the DIP site. Indicates. In addition, primer pairs useful for amplifying the portion of the target polynucleotide containing the SNP or DIP site may be useful, and the amplification product measures the occurrence of nucleotides at the SNP site or the presence of deletions or insertions at the DIP site. Be examined for. Particularly useful methods include those that can be easily adapted to high throughput formats, multiple formats, or both.

  Conditions that allow the generation of an amplification product in the sample in which the amplification reaction is to be performed are such that the reaction includes the components necessary for a possible amplification reaction. Such conditions allow, for example, appropriate buffer capacity and pH, salt concentration, metal ion concentration if required for a specific polymerase, selective hybridization of a primer or primer pair to a template target polynucleotide. Primer or primer extension or amplification product from the appropriate temperature, in addition to polymerase activity and template, or where relevant, forming a secondary structure such as a stem-loop structure Including cycling at a temperature that allows melting of Such conditions and methods for selecting such conditions are routine and well known in the art (eg, Innis et al., “PCR Strategies” (Academic Press 1995); Ausubel et al., “Short Protocols in Molecular Biology”, 4th edition (John Wiley and Sons, 1999), each incorporated herein by reference).

  Primer extension or amplification products can be detected directly or indirectly, and / or can be sequenced using various methods known in the art. Amplification products spanning the SNP site can be used to measure nucleotide occurrence at the SNP locus, 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 chemical degradation methods (Maxam et al., Proc. Natl. Acad. Sci. USA 74: 560, 1977). Can be shinged.

  Nucleotide appearance at the SNP site can also be measured using microsequencing methods, and the identity of only one nucleotide is measured at a given site (US Pat. No. 6,294,336). Microsequencing methods include the Genetic Bit Analysis method (International Publication No. 92/15712). Additional primer-initiated nucleotide incorporation methods for assaying polymorphic sites in DNA have also been described (Komher et al., Nucl. Acids. Res. 17: 7779-7784, 1989; Sokolov, Nucl. Acids Res. 18: 3671, 1990; Syvanen et al., Genomics 8: 684-692, 1990; Prezan et al., Hum. Mutat. 1: 159-164, 1992; Nyren et al., Anal. Biochem. 208: 171-175, 1993). These methods are different from Genetic Bit ™. Analysis in that they all rely on the incorporation of labeled deoxyribonucleotides to distinguish between bases at polymorphic sites. In such a format, the signal is proportional to the number of incorporated deoxyribonucleotides, and polymorphisms that occur in a stretch of the same nucleotide result in a signal that is proportional to the stretch length (Syvanen et al., Amer. Hum. Genet. 52: 46-59, 1993).

  Another method for measuring nucleotide appearance at SNP positions is described by Macevicz (US Pat. No. 5,002,867), where nucleic acid sequences are measured by hybridization with multiple mixtures of oligonucleotide probes. According to such a method, the sequence of the target polynucleotide is determined by allowing the target to hybridize sequentially to a set of probes having an unaltered nucleotide at one position and a different nucleotide at another position. Is done. The nucleotide sequence is determined by hybridizing the target to a set of probes and then measuring the number of sites (i.e. the number of matches) at which at least one member of the set can hybridize to the target. Is done. This process is repeated until each member of the set of probes has been tested. U.S. Pat. A solid phase sequencing method for measuring any one of the sequences is provided.

  The nucleotide appearance of SNPs in a sample can also be measured using the SNP-IT ™ method (Orchid BioSciences, Inc., Princeton, NJ). In general, SNP-IT ™ is a three-step primer extension reaction. In the first step, the target polynucleotide is isolated from the sample by hybridization to the capture primer, giving a first level of specificity. In the second step, the capture primer is extended from a terminating nucleotide triphosphate at the target SNP site, giving a second level of specificity. In the third step, extended nucleotide triphosphates can be detected using a variety of known formats including: direct fluorescence, indirect fluorescence, indirect colorimetric assay, mass spectrometry, Such as fluorescence polarization. The reaction can be processed in an automated format in a 384 well format using a SNPstream ™ instrument (Orchid BioSciences, Inc., Princeton, NJ). Phase known data can be created by inputting phase unknown raw data from the SNPstream ™ device into the Stephens and Donnelly's PHASE program.

  Melting curve analysis of SNPs (McSNP® analysis) provides another method for detecting nucleotide occurrences in AIM (Akey et al., Supra, 2001). McSNP® analysis does not require a gel electrophoresis step, thus minimizing the time and expense to detect SNPs, and can be easily adapted to high throughput formats, and thus Providing the additional advantage of allowing parallel examination of one or more AIM panels and / or samples.

  Where a particular nucleotide occurrence of a SNP is such that the nucleotide occurrence results in an amino acid change in the encoded polypeptide, the nucleotide occurrence is indirectly identified by detecting a particular amino acid in the polypeptide. Can be done. The method for measuring amino acids depends, for example, on the structure of the polypeptide or on the position of the amino acid in the polypeptide. Where a polypeptide contains only a single occurrence of an amino acid encoded by a particular SNP, the polypeptide can be examined for the presence or absence of that amino acid. For example, simple sequencing of terminal amino acids can be performed where amino acids are at or near the amino terminus or carboxy terminus of a polypeptide. Alternatively, the polypeptide can be treated with one or more enzymes, and a peptide fragment containing the amino acid position of interest detects the specific movement of the peptide, for example, by sequencing the peptide or after electrophoresis Can be examined. Where a particular amino acid contains an epitope of a polypeptide, specific binding of an antibody specific for the epitope, or its absence can be detected. Other methods for detecting a specific amino acid in a polypeptide or peptide fragment thereof are well known and include, for example, the convenience of a device such as a mass spectrometer, a capillary electrophoresis system, a magnetic resonance imaging device, etc. It can be selected based on effectiveness.

  In another embodiment, the method of the invention specifically binds to a polypeptide comprising an amino acid encoded by a nucleotide sequence comprising, for example, one nucleotide occurrence of a SNP, but is encoded by a codon comprising that SNP. Does not substantially bind to polypeptides comprising different amino acids; or binds specifically to a polypeptide comprising an amino acid sequence encoded by, for example, one type of DIP (eg, with an insertion), but an alternative type An antibody or antigen-binding fragment thereof that does not substantially bind to that encoded by (eg, has a deletion) is utilized. As used herein, the term “specific interaction” or “specifically binds” means that two molecules form a complex that is relatively stable under physiological conditions. To do. The term is, for example, the interaction of an antibody that binds a target polynucleotide containing a SNP site only if the SNP has been identified but not substituted, only if it has a nucleotide occurrence (eg, A, but not T); or A variety of interactions, including antibody interactions that bind a polypeptide that contains a single amino acid encoded by a codon that includes the SNP site, but do not bind a polypeptide that has an alternative amino acid encoded by the codon that includes the SNP. Used herein to refer to action.

Specific interactions are at least about 1 x 10 ^-6 M, generally at least about 1 x 10 ^-7 M, usually at least about 1 x 10 ^-8 M, and especially at least about 1 x 10 ^-9 M or 1 It can be characterized by a dissociation constant of x 10 ^-10 M or higher. A specific interaction is generally a condition occurring in a living individual such as, for example, a human or other vertebrate or invertebrate, as well as a mammalian cell or another vertebrate organism or invertebrate It is stable under physiological conditions, including those occurring in cell cultures such as those used to maintain cells from animal organisms. Methods for determining whether two molecules are interacting specifically are well known and include, for example, equilibrium dialysis, surface plasmon resonance, and the like.

  Antibodies useful in the methods of the present invention bind specifically to a polynucleotide comprising AIM, or to a polypeptide comprising an amino acid encoded by a codon comprising SNP or comprising an amino acid by insertion at a DIP site including. Such an antibody specifically binds a polypeptide comprising a first amino acid encoded by a codon comprising a SNP locus, but comprises a second amino acid encoded by a codon comprising a different nucleotide occurrence in the SNP. Are selected so as not to bind or to be measurable weaker.

  The term “antibody” is used broadly herein to refer to immunoglobulin molecules that specifically bind antigen and antigen-binding portions of immunoglobulin molecules. As such, antibodies useful in the methods of the present invention include polyclonal, monoclonal, multispecific, human, humanized or chimeric antibodies, single chain antibodies, Fab fragments, F (ab ′) fragments, Fab expression libraries. Fragments produced, anti-idiotype (anti-Id) antibodies, etc., plus 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 (scFv), single chain antibody, disulfide bond Fv fragment (sdFv) and VL or Includes fragments containing any of the VH domains. Thus, an antigen-binding antibody fragment comprising a single chain antibody can comprise a variable region alone or in combination with all or part of the hinge region, CH1, CH2 and / or CH3 domain. The antibody can be from any animal origin, including birds and mammals, or can be recombinantly expressed, eg, in insect or mammalian host cells, or in plants.

  There is much that can be learned today through the use of genetic markers in many scientific fields. Although the use of gene sequences has become routine for forensic and disease research, most of the benefits from the recently completed human genome project are still awaiting discovery. Within the genome are our quality of life through increasing crop yields, extending human life expectancy, minimizing drug-induced pain, and better, more effective and specific treatments There are sequences and patterns of sequences that will prove useful for a variety of purposes, including improving To date, biomedical research has been conducted in relatively simple terms. Nevertheless, more than 1,000 simple Mendelian traits were mapped by following the transmission of genetic markers in the family.

  Many statistical methods are available to study genetic traits, including traditional pedigree-based linkage analysis, distributed component methods, sibling pair linkage, measurement genotype, transmission disequilibrium, genome control and structural analysis is there. Some of the genes that underlie variability in susceptibility to common diseases (eg, heart disease, obesity, type 2 diabetes, hypertension and cancer) will eventually be identified using genetic approaches. However, in genetic studies in certain diseases, many of these conditions are multifactorial (i.e., have some cause of variability in risk) and multigenic (i.e. effects between several genes). And the result of interaction), there is a great deal of complexity. Further difficulties in the study of common diseases can stem from late onset of symptoms and heterogeneity in etiology. Thus, identifying genes involved in complex diseases remains one of the best challenges in the field of human genetics.

  Increasing interest in association analysis as a useful approach to map common diseases and drug response genes (Risch and Merikangas, Science 273: 1516-1517, 1996; Jorde, Genome Res. 10: 1435-1444, 2000; Nordborg and Tavare, Trends Genet. 18: 83-90, 2002). However, until the present disclosure, the ancestry implications for identifying these genes were not fully recognized. As such, the methods of the present invention provide a platform not previously described 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 for inferring an individual's response to a commonly used drug, which is notably a function of the individual's ancestry, are provided; the disclosed markers and methods differ to varying degrees for each drug Until useful in the inference of such responses. Further provided are compositions and methods for inferring the ancestry of an individual and / or group from knowledge of the DNA sequence of the individual or group. Still further, compositions and methods are provided for using knowledge of ancestry-related DNA sequences to identify disease susceptibility and drug response genes by the MALD process. Also provided are compositions and methods for authorizing and standardizing research groups for more traditional methods of mapping disease genes. Each of these processes requires accurate knowledge of ancestry that can be measured using the methods and compositions disclosed herein.

  The population that seemed best suited for linkage disequilibrium (LD) mapping prompted much discussion and discussion (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., Amer. J. Hum. Genet. 70: 673-685, 2002). The degree of LD is a complex function of a number of genetic and evolutionary factors such as mutation, recombination and gene conversion rates, demographic and episodic events, and the age of the mutation itself. Some of these factors affect the entire genome, while others only affect specific genomic regions. Furthermore, variations in the rate of mutation, recombination and gene conversion throughout the genome are expected to cause LD differences between genomic regions (e.g., Tailon-Miller et al., Nat. Genet. 25: 324-328, 2000 ).

  It has been proposed that small, isolated, inbred populations will have advantages over other populations due to lower heterogeneity and a greater degree of linkage disequilibrium (Wright et al., Supra, 1999; Nordborg and Tavare, supra, 2002; Kaessmann et al., Supra, 2002). Other populations that are well suited for mapping are recently mixed populations (eg, Hispanic and African Americans), providing the advantage that LD was recently caused by the mixing process. Because this LD is recent, it can span large chromosomal regions. However, it is also extremely important to control the genetic structure (individual variability in the mixing ratio) present in these populations to avoid false positives (Parra et al., Supra, 1998; Lautenberger et al., Amer. J. Hum. Genet. 66: 969-978, 2000; Pfaff et al., Amer. J. Hum. Genet. 68: 198-207, 2001; Nordborg and Tavare, supra, 2002, each incorporated herein by reference. ) Interest in mixed mapping has increased in recent years (McKeigue et al., Ann. Hum. Genet. 64: 171-186, 2000; Smith et al., J. Invest. Dermatol. 111: 119-122, 2001; Collins-Schramm et al. Amer. J. Hum. Genet. 70: 737-750, 2002, each incorporated herein by reference). A general description of mixed mapping is a statistical approach developed for mixed mapping and some items about its application to skin pigmentation as a model phenotype, but is provided below.

  Mixing results in an allelic association between all marker loci that differ in allelic frequency between parental populations (Chakraborty and Weiss, Proc. Natl. Acad. Sci., USA 85: 9119-9123, 1988). These associations fade with time in a manner that depends on the genetic distance between them. In this way, disease (or trait) risk alleles that differ between parental populations can be mapped in a mixed population using a particular panel of genetic markers that show a high frequency difference between parental populations. These markers, called AIM, are characterized by having certain alleles that are more common in one group of populations than in other populations. One measure of the informability of such markers is the allelic frequency difference, delta (δ), which is simply the absolute value of a particular allelic difference between populations (Chakraborty and Weiss, supra, 1988; Dean et al., Supra, 1994).

  In mixed populations, allelic associations have recently occurred and are therefore given because they span longer distances (up to 10-20 centimorgans (cM) or more) in unmixed populations. The detected sample size is more easily detected. The statistical basis of this approach was first by Chakraborty and Weiss (supra, 1988), and then the method was “mapped by mixed linkage disequilibrium” (MALD; Stephens et al., Amer. J. Hum. Genet. 55: 809. -824, 1994; Briscoe et al., J. Hered. 85: 59-63, 1994), explored by Stephens, Briscoe and O'Brien. In addition, individuals excluded from marker data, whether to use the MALD approach for genetic studies or the more traditional LD approach to eliminate associations between alleles and traits at unlinked loci Need to be controlled in the analysis of the ancestors. The SNP sequences (markers; AIM) and methods (BGA testing) disclosed herein are particularly efficient means to accomplish this task. An analysis of covariance (ANCOVA) test was used, and an estimate of individual mixture was used as a conditioning variable to control for the influence of individual ancestry in the following two ways: 1) Exclude the locus under consideration ( ANCOVA / IAE minus marker); and 2) Use full individual ancestry estimates for conditioning (ANCOVA / IAE). This method is described in detail herein.

  An alternative approach to exploring mixing has been developed that is based on earlier studies, but has little in common with classical LD mapping and is more similar to experimental cross-linkage analysis (McKeigue Amer. J. Hum. Genet. 63: 241-251, 1998, incorporated herein by reference; McKeigue et al., Supra, 2000). For this reason, the term “mixed mapping” has been proposed as more appropriate than “mapped by mixed linkage disequilibrium”. Instead of testing for allelic association, the method models the underlying variation in ancestry onto a mixed pedigree chromosome to extract all information about linkages resulting from mixing. The disclosed methods and markers are necessary and sufficient to accomplish this process. Although the underlying principle that relies on detecting linkages is straightforward, advanced statistical methods are used to apply this method in practice. For example, assume that the locus is responsible for some of the variations in pigmentation between West Africans and Europeans. Individuals of mixed pedigrees kept other factors constant when classified according to whether they had 0, 1, or 2 African ancestry alleles at this locus In comparing these three groups, the average pigmentation level appears to vary with the percentage of alleles at loci that are of African ancestry. Controlling the analysis for parental mixing removes the trait association with ancestors at unlinked loci and ensures that other factors remain constant and comparisons are made.

  To infer the allelic ancestry at the locus from the marker genotype, the conditional probability of each allelic state is assumed to be the ancestor of the allele (ancestor-specific allelic frequency), for example, West Africans or Europeans Needed if you want. There is increasing evidence that mixed mapping is an effective means of gene identification, and in mixed populations, strong allele associations are observed between linked markers spaced at substantial distances (Parra et al., Supra, 1998; Parra et al., Amer. J. Phys. Anthropol. 114: 18-29, 2001; McKeigue et al., Supra, 2000; Lautenberger et al., Supra, 2000; Smith et al., Supra, 2001; Wilson and Goldstein, Amer. J. Hum. Genet. 67: 926-935, 2000; Pfaff et al., Supra, 2001). If very high levels of association were observed over long genetic distances, phenotypes that differ between parental populations due to several genetic factors are also expected to show association with linked AIM Is done. A well-suited phenotype for applying mixed mapping is skin pigmentation.

  Despite the power of AIM for disease genes and forensic analysis, no studies have been conducted to elucidate this power. As disclosed herein, 1) SNPs or deletion / insertion polymorphisms (collectively referred to as AIM) in the human genome that may be useful for drug response, disease genes or forensic research 2) biochemical and genetic test results are provided that demonstrate that these AIMs may be useful for disease genes and forensic research; 3) actual drug response, disease genes or Demonstrates the usefulness of AIM derived from a systematic screen of the human genome in forensic research; 4) From a systematic screen of the human genome to infer whether individuals are likely to acquire disease or are less responsive to drugs The usefulness of the extracted AIM is demonstrated; 5) Crime scene DNA specimens, for example 80% European, 10% African and 10% Asian heritage or some other Demonstrates the usefulness of AIM derived from the systematic screen of the human genome to infer whether it was from a ratio / mixed individual; 6) infer the percentage of ancestors of individuals from those DNA The utility of AIM derived from a systematic screen of the human genome for whether the individual is 80% European, 10% African and 10% Asian heritage or some other ratio / mix; And 7) infer the proportion of ancestors of the group of individuals from their DNA (e.g., the group can be a population sample, a family or a group of clinically defined people, but 80% European, 10% The utility of AIM derived from a systematic screen of the human genome (whether African and 10% Asian heritage or some other ratio / mix) is demonstrated.

  The results show that AIM is useful for the above applications, and that the sequences exemplified herein and additional AIMs identified using the methods disclosed herein allow these applications Prove that. The AIMs and methods of the present invention are useful for the study of human disease, drug response and physical traits, and therefore provide exceptional commercial potential. For example, during this stage of personalized prescription and disease risk assessment, the markers and methods of the present invention provide the tools needed to advance in this pioneering industry. As illustrated herein, an individual's response to a particular drug is determined by the individual population in addition to, but irrespective of, the person's genotype or xenobiotic metabolism gene sequence for the drug target. Dependent on the degree of structure (ie, the heritage of a particular ancestor). As such, the compositions and methods of the present invention provide a means for predicting the likelihood of an individual to respond to a particular drug.

  For example, in a screen of genetic markers related to patient response to a cholesterol-lowering drug, Lipitor ™, identified for LDL response to Lipitor ™ by a low-density lipoprotein (LDL) response, a favorable response indicator Some of the most potent markers that were made were gene types that were not immediately recognized as relevant for drug response, including, for example, TYR, OCA2, TYRP, FDPS, and HMGCR (WO 03 See also / 002721 (PCT / US02 / 20847) and WO03 / 045227 (PCT / US02 / 38345), each incorporated herein by reference). When combined with markers from genes that are biologically relevant for the response, they increase the ability to derive an accurate inference of the response from DNA. Each of these markers is an excellent AIM, indicating that the AIM linkage to drug response is likely a function of ancestral differences in response trends (see Example 5). As such, the ancestor heritage may predict a favorable response to Lipitor ™. This association has been observed for almost any type of response (n = 54) to almost any type of drug examined (n = 23), thus inference of drug response is at least partially inferred from ancestral proportions Confirmed that it can be achieved through As such, genes that are truly relevant for drug response are at least partially a function of the ancestry of the individual, and gene sequences that are relevant for drug response are markers that provide information about ancestry (i.e., AIM ) And seem to be statistically linked.

  Screening the genome for the true identification of genes associated with specific traits such as drug responsiveness is prohibitively expensive and time consuming. As such, the use of AIM to infer an individual's propensity to drugs is for the rapid development of tests that can be used to find those drugs that fit patients and are most appropriate for their genetic makeup. Provide meaningful shortcuts. Thus, in addition to being useful for mixed mapping of disease genes, the disclosed methods and illustrated markers provide a tool that can direct treatment protocols by clinicians. The ability to use AIM effectively for the identification of AIM from publicly available human genome data, and for the development of patient-drug classification sets, mixed screening panels and forensic tools, is the SNP database for AIM (e.g., URL `` Screening the World Wide Web (see “www”) at nih.ncbi.nlm.gov); screening AIM against a multi-ancestral panel of DNA samples to demonstrate what is truly a good AIM Using the disclosed statistical and software methods for using AIM sequences to guide biologically relevant inferences; and the likelihood of an individual responding to a drug or developing a disease Achieved using the disclosed methods, including recognizing what can be expected through knowledge of their ancestors, then shown through the AIM sequence of the individual .

  Prior to this disclosure, individual ancestry could be estimated using two independent methods: maximum likelihood (Hanis et al., Amer. J. Phys. Anthropol. 70 (4): 433-441, 1986, as a reference. And the Bayesian method implemented in the STRUCTURE program (Pritchard et al., Genetics 155: 945-959, 2000, incorporated herein by reference). Although the maximum likelihood and Bayesian methods provide proportional ancestry or mixed point estimation, there are several deficiencies in these methods that are addressed by the disclosed methods. For example, when using the disclosed algorithm (see Example 6; see also Table 12 containing a flowchart illustrating the algorithm), 1) the most likely group from which the individuals are derived is the proportional ancestor estimation Estimated at the same time; 2) the multidimensional confidence interval was computed and projected by the computer, thus reducing the complexity of presentation; 3) each level in the past (parent, grandparent, great-grandparent , Etc.) has been developed to estimate the number of ancestors and their proportions mixed; and 4) the proportional BGA affiliation within an individual for more than two BGA groups is derived at one time, and thus, for example, A classifier for providing improved and more accurate forensic applications, in addition to quantitatively or continuously distributed traits (i.e. not dichotomous), trait values are at least partly a function of BGA Opening It made it possible to.

  Independent methods for the classification of individuals into population groups have been developed (Shriver et al., Supra, 1997, Frudakis et al., Supra, 2002, each incorporated herein by reference). This method differs from previous classification methods in that it allows the simultaneous estimation of the optimal group to which a particular individual belongs and the proportional assignment of individuals to multiple parent groups. Example 6; see also Table 12). Thus, where the previous method has made it possible to declare that a person is much more likely to be African American than European American, this approach allows the same statement, And also provides the individual's proportional ancestry with a confidence interval (CI); for example, 25% (95% CI 15-35%) European ancestry; 75% (95% CI 60-80%) African ancestry; And 0% (95% CI 0-6%) Native American ancestry. In addition, confidence intervals can be represented in multidimensional space to provide a clearer representation of measured ancestry for the person in question (see below; see also FIG. 2). Although methods for drawing such displays are known, the present disclosure is the first to provide a display that is presented with quantifiable confidence.

  There is a clear difference in chromosome segment ancestry patterns (PCSA) between persons with different ancestor history (see Figure 1). A series of AIMs across chromosomes can facilitate the estimation of the most likely parental combination leading to the sequence profile observed in a given person. One example of where PCSA estimation is important is in the identification of Hispanic ancestors from people with primarily European ancestry, including some proportion of recent Native American ancestry. In fact, this is an important measurement because the political and legal rights required and granted to these two groups can depend on their ancestry. Hispanic groups such as Mexican Americans (MA) have approximately 30-40% Native American ancestry, but the balance is European with a small fraction (about 5%) of African ancestry. A person who is a quarter indigenous American has 25% indigenous American ancestry and therefore seems to overlap with many MAs at his level of estimated ancestry. The PCSA pattern is significantly different for these two cases, and in such cases it is expected that it can provide some of the only genetic evidence that facilitates the precise definition of ancestry. As disclosed herein, PCSA can be used in ancestral studies.

  An important step in these measurements is the phasing of AIM along the chromosome segment (see Example 2, FIG. 8; see also Example 5, FIGS. 12-16). Matching AIM phases along the chromosome is: 1) inferring from individual genotype, 2) molecular haplotyping (e.g., allele-specific PCR combined with genotyping), and 3) single sperm analysis (For female subjects, male sibling sibling sperm would give the same profile) and can be achieved in several ways. Furthermore, the disclosed method allows for simultaneous consideration of two sex chromosomes (X and Y) and mtDNA for ancestral reasoning. AIM can be found in each of these sources and can inform many of the problems related to the proportion of human ancestry and the population from which a particular person is derived. For example, the Hispanic / Latin American population has a very high (65-100%) frequency of Native American mtDNA haplogroups, but shows only a small contribution from the Native American population in autosomal markers. Thus, for example, a person with an ancestor who is said to be an indigenous American on her father's side who has a non-indigenous American mtDNA haplogroup is partly more indigenous than an indigenous American. It's more likely not to be Hispanic than she was destined to have an Native American mtDNA haplogroup.

  Linkage disequilibrium (LD) is increasingly being used as a mapping tool for both detailed measurement of gene location and initial localization of disease genes in specific populations. Allelic associations are significantly non-random and correlated with physical distance within a small (<60 kb) genomic region (for review, see Jorde, Amer. J. Hum. Genet. 66 : 979-988, 1995; see Jorde, Genome Res. 10: 1435-1444, 2000), probably reflecting the underlying "block structure" that characterizes many genomic regions (Reich et al., 2001; Daly et al., 2001). Thus, when disease alleles in a population share a recent common origin, the closest genetic marker with the strongest association is likely the closest to the locus causing the disease. This approach was important in the positional cloning of several simple Mendelian diseases, including the cystic fibrosis gene, the Huntington's disease gene, and the crustal dysplasia gene.

  In addition to application in detailed mapping or positional cloning, it can be used for initial disease gene mapping in a homogeneous population that has recently increased in size or is genetically inbred. In such populations, disease alleles are probably present in a few founders, and recombination has limited opportunities to randomize associations between these alleles and linked marker loci. It was. Analysis of allelic associations between affected and unaffected individuals from these populations may thus facilitate localization of disease loci. A number of Mendelian diseases have been mapped using this approach: several diseases in the Finish population, Hischsprung's disease in Menonite, benign recurrent intrahepatic cholestasis in an isolated Dutch fishery community Stagnation, early-onset familial persistent hyperinsulinemia in Saudi Arabia, and Bardet-Biedl syndrome in Bedouin.

  There has been a lot of controversy as to which population is best suited for LD mapping of complex polygenic diseases (eg, Wright et al., Supra, 1999; Eaves et al., Supra, 2000; Nordborg and Tavare, supra, 2002; (See Kaessmann et al., Supra, 2002). The degree of LD is a complex function of a number of genetic and evolutionary factors such as mutation, recombination and gene conversion rates, demographic and episodic events, and the age of the mutation itself. Some of these factors affect the entire genome, while others only affect specific genomic regions. Furthermore, variations in the rate of mutation, recombination and gene conversion throughout the genome are expected to cause LD differences between genomic regions.

  Small, isolated, inbred populations outperform other populations with lower heterogeneity and greater degree of linkage disequilibrium in terms of populations for use in disease discovery attempts performed by LD-based methods It has been proposed to have advantages (see Wright et al., Supra, 1999; Nordborg and Tavare, supra, 2002; Kaessmann et al., Supra, 2002). On the other hand, mixed populations such as Hispanics and African-Americans offer the advantage that linkage disequilibrium has recently been caused by the mixing process and it can span large chromosomal regions, but avoids false positives It is very important to control for the genetic structure present in these populations (Parra et al., Supra, 1998; Lautenberger et al., Supra, 2000; Pfaff et al., Supra, 2001; Nordborg and Tavare, supra, 2002) . Despite increasing research focus on LD-based methods, however, many issues regarding LD in the human population remain unexplored. Currently, NHGRI is planning a systematic project to help develop information tools for gene identification studies by identifying common haplotypes in several populations. This “Haplotype Map Project” (HMP) is likely to be an attempt across multiple large institutions focused on finding common haplotypes in common population samples. HMP is likely to prove to be an important data source for identifying AIMs as disclosed herein, as some populations are examined for haplotype block structure, and therefore additional candidates Provides a basic plan for detailed LD structure in AIM and part of the parental population.

  A mixed mapping as disclosed herein is complementary to HMP, although it differs in nature from HMP. First, the primary focus of HMP is to understand the detailed structure of individual genomic regions in the genome, but this method allows an understanding of LD that is specifically caused by mixing. The level of LD from mixing is on the order of millions of bases (Mb; megabase) and tens of Mb, while HMP focuses on tens of kilobases (kb), The genomic and population characteristics that affect the results from the project may not be noticed elsewhere. Second, mixed mapping requires accurate parental allele frequency estimation. As such, a number of different African, Native American, European and Asian populations were typed (see Table 6 below), but HMP focuses on one or two samples of the main population groups. There is a high possibility of matching.

  Third, a large number of African-American and Hispanic samples (n = 500) are typed and have sufficient statistical power to test the coverage of mixed maps and compare analytical methods accordingly. provide. In addition, as several representative populations from different regions of the country have been typed, geographic variations in ancestral proportions and mixed dynamics can be examined. Although some mixed populations are likely to be included in the HMP, the number of individuals and the number of different population samples are less than those disclosed herein and therefore the same comparison is not possible . For example, having 10 samples for each of four ancestry groups is not sufficient for identification of sequences preferentially present in one or several of those groups; as disclosed herein At least 50 individuals were examined for each of several tens of ancestry groups (not just 4) to comprehensively identify these markers.

  Fourth, recent population variability attempts (e.g., SNP consortium allele frequency projects) and perhaps the focus of HMP is to exclude indigenous American populations for many complex reasons, East Asians, Africa Was in the human and European samples. However, the exclusion of these populations is in understanding the genetic nature of the fastest growing group of US inhabitants, namely Hispanic, having a significant level of indigenous American ancestry (20% -40%). Cause shortages. With the markers and methods disclosed herein, the genetic nature of a Hispanic population's disease can be examined. Similarly, several different indigenous American populations may represent important parental populations for many distinct groups that are often classified together as Hispanic.

  The population-based related methods disclosed herein provide several advantages over traditional linkage studies. Estimating the location of disease genes by traditional genetic linkage methods relies on the use of either a related person, an extended multi-generation family or a pair of related individuals. These approaches are effective and very powerful when investigating diseases caused by a single gene. However, polygenic and multifactorial diseases, such as type 2 diabetes, hypertension and prostate cancer, arise from the interaction of several genes and complex environmental effects, and can be studied using traditional methods Is more difficult. The identification of genes that contribute to susceptibility to common diseases is complicated by heterogeneity. The source of genetic heterogeneity determines which mapping method is most likely to work for gene identification. There are two basic types of genetic heterogeneity: locus heterogeneity, where more than one locus affects the genetic trait, and phenotype within the locus that is the specific cause There is allelic heterogeneity, where there are multiple alleles that are important in changing. Traditional linkage analysis using extended families generally does not respond to allelic heterogeneity, but can be adversely affected by locus heterogeneity. Alternatively, LD-based methods are generally adversely affected by allelic heterogeneity, but little affected by locus heterogeneity.

Given little allelic heterogeneity, association-based approaches such as measured genotypes and transmission disequilibrium tests (TDT) can be more sensitive than family-based LOD scores or sibling pair methods. Risch and Merikangas (supra, 1996) compare the number of individuals required for sibling pair studies and TDT studies, and the number of individuals required to detect linkage is more for TDT than for sibling pair studies. It was very small. This is especially true when the effect of the disease locus is small. For example, for a locus with a risk rate of 2.0 and a gene frequency of 50%, 2500 sibling pairs or 340 cases / parents for TDT are required. There are several examples where the related demonstration using Haplotype Relative Risk (HRR) or case / control design, or linkage in TDT, has demonstrated linkage in sibling pairs. A classic example is the association between the insulin gene and IDDM, which was demonstrated in cases and controls and subsequently confirmed using TDT, but was often not observed in sibling pair linkage studies (Spielman et al., Amer. J. Hum. Genet. 28: 317-331, 1993). Yaouanq et al. (Science, 1997) are very significant about the linkage between HLA and multiple sclerosis using TDT in a series of 157 French families (99 single and 58 multiple) ^N (p <10 ^-9 ) evidence was reported. When 58 multiple families were analyzed alone, p-values of 0.0001 and 0.03 were generated for the TDT and sibling pair methods, respectively.

  Association analysis based on candidate genes is relatively more powerful than family linkage analysis to detect disease genes, but a thorough examination of all genes in genomes with more than 40,000 genes is currently Not practical. The haplotype map project may succeed in creating the necessary information sources for gene identification based on linkage disequilibrium. However, even though the block structure model of the human genome can be explained by four haplotypes in each gene, the minimum number of SNPs and DIPs is 80,000, and the actual number may be higher. Although genotyping technology is advancing rapidly, testing this number of markers in many research subjects is not yet practical. In addition, there are several important assumptions inherent in the plan to identify genes using LD in large populations. One important difficulty in using linkage disequilibrium in genome-wide screening is that LD declines exponentially with the recombinant fraction between the marker and disease locus, and with the number of mutations that cause the disease It is to be. For older mutations that predispose to disease, LD becomes very weak, even between disease alleles and alleles at relatively closely spaced marker loci.

  LD mapping is a mapping of diseases in rare genetic diseases such as cystic fibrosis and in certain populations such as Finns and Bedouins, significant population bottlenecks, inbred or founder-sensitive populations Was useful. In these situations, the mutant allele is relatively young, as in cystic fibrosis, or the population is present in order to reduce genetic variability and increase LD across the genome. Reading models for the genetic nature of common diseases define alleles that predispose at multiple loci that increase individual risk when present in specific combinations (Greenberg, Amer. J. Hum. Genet. 52: 135-143, 1993; Lander and Schork, Science 265: 2037-2048, 1994; Risch and Merikangas, supra, 1996). If the disease is common, the predisposing allele for this model is also expected to be relatively frequent. However, given the natural model, the frequency of alleles in a population is on average related to the age of the allele, such that more frequent alleles are older than rare alleles. In a homogeneous population, LD is inversely related to the age of the allele, and risk alleles for common diseases are expected to be relatively old on average, so in unisolated or inbred This fact poses a problem for the application of LD-based methods to identify common disease genes in no population.

  Application of the compositions and methods of the present invention for mixed mapping allows for more accurate and reliable mapping of complex traits. Mixed mapping can take advantage of LD caused when previously segregated populations are mixed and avoid these problems in mapping complex traits. It was first recognized by Chakraborty and Weiss (supra, 1988) that mixed populations could be useful in measuring genetic linkage. When genetically diverse populations hybridize, nonrandom allele associations occur between loci with significant allelic frequency differences, even between unlinked loci. This LD declines rapidly if the genetic locus in question is not located close together on the same chromosome.

LD declines as a function of the recombination rate between two markers (θ) and the number of generations from those hybridizations (n), where D _n = (1-θ) ⁿ D ₀ , D _n is the hybridization The latter n generations of linkage disequilibrium and D ₀ can be expressed as being the first linkage disequilibrium (Chakraborty and Weiss, supra, 1988). Assuming this exponential relationship between the decrease in LD and genetic distance, a mixed population in which markers are close together and remain high because of genetic linkage (if the time from mixing is short) ) And background linkage disequilibrium between unlinked loci. For example, after 10 generations, linkage disequilibrium at unlinked loci drops to 0.1% of the initial level, and at 10 cM and 1 cM away loci, true linkage disequilibrium is still Of 34.9% and 90.4%, respectively. A critical parameter for effective detection of linkages in the identified mixed population is the frequency difference (δ) between the parent population and the number of generations from hybridization. Linkage by association analysis in mixed populations works efficiently when δ is large (greater than 0.2) and the number of generations from mixing is small (in the order of 10 generations; Chakraborty and Weiss, supra, 1988). It was.

  Stephens et al. (Supra, 1994) and Briscoe et al. (Supra, 1994) studied this approach using computer simulation (MALD) and detailed practical considerations for study design. Using a simple model of the mixed type that occurred in the United States, they were typed on 200-300 equally spaced markers, each of 200-300 patients with δ> 0.3. Using the sample size suggested that there would be> 95% opportunity to locate the causative gene. A consistent result from several models studied was the primary dependency of MALD power on the allelic frequency difference of the markers used. When δ is small, the initial LD is small and difficult to distinguish from background noise.

  Stephens et al. (Supra, 1994) suggested using loci where δ> 0.4 between mixed parental populations for effective mixed mapping. They also found that mixed mapping was most effective in populations that hybridized 4-20 generations ago, and that there was no incremental gene transfer (provided that there were no new introgressions from the parental population during the last three generations). Slow gene transfer from one population to another; also known as a continuous gene flow model) has been demonstrated to affect the power of mixed mapping, although not critical. The disclosed mixed mapping technique can identify the location of disease susceptibility genes by analysis of a mixed population composed of parental populations with large differences in the frequency of susceptible genotypes. As such, mixed mapping applications include studies of type 2 diabetes susceptibility in Pacific Islander populations, hypertension, obesity and prostate cancer in African Americans, and type 2 diabetes, obesity and gallbladder disease in Hispanic populations .

  McKeigue has developed an approach to explore mixing in mapping genes based on earlier work (McKeigue, supra, 1997; McKeigue, supra, 1998; McKeigue et al., Supra, 2000). The approach is powered by LD generated by mixing, but is more similar to linkage analysis of experimental crossovers. For this reason, the term “mixed mapping” has been proposed. Instead of testing for allelic association, a model of the underlying variation in ancestry can be created on the mixed pedigree chromosomes to extract all the information about linkages resulting from mixing.

  As discussed above, advanced statistical methods are needed to apply this approach in practice. Conditioning in the parental mix removes the association of the trait ancestors at unlinked loci and ensures that other factors remain constant and the comparison is made. In non-statistical terms, each individual is compared for the proportion of alleles at a marker locus that is of a particular pedigree with the expected proportion, assuming a mixture of the individuals' parents. One simple way to do this is to use a covariance analysis (ANCOVA) test (see Tables 2 and 3), but this simpler approach may not use all of the available information. Absent. As such, the Bayesian method was also used (see Tables 2 and 3).

  An ancestor-specific allele frequency is required to infer the allelic ancestry at the locus from the marker genotype; ie, given the allelic ancestry (in this example West African or European) Conditional probability for each allelic state. The entire set of alleles at any locus in a mixed population can be considered to be composed of two subpopulations-alleles of African ancestry and alleles of European ancestry. As long as the ancestor-specific allele frequencies are correctly specified for the mixed population under study, Bayes' theorem reverses these conditional probabilities, and for each individual under study, the next distribution of ancestors at the locus It can be applied to calculate (0, 1 or 2 alleles of African ancestry). If the information conveyed by typing a single marker is not sufficient to assign each allelic ancestor at the marker locus to one of the two founding populations, the marker is ancestor at the adjacent locus. Can be combined with multipoint analysis to estimate.

  Simulation studies have shown that with sufficient markers, a high percentage of information about ancestry at each locus can be extracted even if no single marker provides complete information about ancestry ( McKeigue, supra, 1998). Based on these simulations, a panel of markers with Fst> 0.4 at an average interval of 2-3 cM for all 1,000 AIMs can be constructed as disclosed herein. It should be appreciated that a panel of 1,000 AIMs for a particular population (eg, a group of African Americans, primarily West Africans and Europeans) often overlaps with panels for other groups. In other words, the AIM selected for one level of identification (eg, African / European) often also provides information about other identifications (eg, Native American / European). Table 1 lists the first identified panel containing 32 AIMs (SEQ ID NOs: 332-363; see also Example 1). Using a cut-off of d> 0.3, only four of these markers were transferred to one of three comparisons (African / European; African / Indigenous American; Native American / European) Limited in providing information; the rest gives information for two of the comparisons, and one marker gives information for all three comparisons. In further studies, a panel of 71 AIMs were identified as informative about Indo-Europeans, Sub-Saharan Africans, Native Americans and East Indians (SEQ ID NOs: 1-71; Table 6) (Examples) 2).

  There is increasing evidence that mixed mapping is an effective means of gene identification. At least three independent groups reported strong mixed linkage disequilibrium (ALD) between linked markers spaced at a substantial distance (eg, Parra et al., Supra, 1998 and 2001; Pfaff et al., Supra). 2001; McKeigue et al., Supra, 2000). Given the very high level of association observed over long genetic distances, phenotypes that differ dramatically between parental populations due to some genetic differences are also linked to AIM It is expected to show the relationship. However, until this disclosure, no systematic screen has been reported to identify SNPs based on the required version of AIM that the MALD approach is likely to show. McKeigue and others have identified a panel of STR AIMs for use in this approach, but the use of STRs for this purpose is to accurately estimate STR allele complexity and allele frequency. There is a problem because of the large database required. Even small errors or even incomplete assumptions in the frequency of alleles that have not been observed can amplify the statistical power of the study.

  Inhomogeneities within the parental population can have a disruptive effect on mixed mapping studies. In the case of African-American populations, the mixing process that took place in the New World included mainly heterogeneous groups of populations from Central West Africa and Europe, plus several indigenous American populations. . With regard to the genetic contribution of Europe, the most important source populations came from Great Britain, Ireland, Germany and Italy. It is important to show the relative homogeneity of the European population from a genetic point of view despite the diverse geographical regions of origin of the parent European population (e.g. Cavalli-Sforza et al., Supra, 1994) .

  With respect to African contributions, it is well known that the African continent contains a tremendous amount of genetic diversity. However, only a subset of African genetic diversity contributed to the formation of African-American populations. The vast majority of slaved Africans came from Central Western Africa, approximately from northern Senegal to southern Angola (Curtin, The Atlantic Slave Trade; Madison, University of Wisconsin Press 1969); other parts of Africa Was not affected by the slave trade. Of the four main linguistic languages in Africa, the Niger-Congo Cordofan, the Nile-Sahara, the Afro-Asian and the Koisan (Greenberg, supra, 1963) are forced into the New World. The majority of slave Africans were Niger-Congo members. This widespread language includes West African (spoken to people from Senegal to Nigeria) and Bantu (dominant in Central and South Africa). The Bantu language is dispersed throughout Africa due to the “recent” expansion that occurred about 3,000 years ago, perhaps in West Africa (Nigeria and Cameroon; Excoffier et al., Yearbook Phys. Anthropol. 30: 151-194, 1987 and Cavalli-Sforza et al. , Supra, see 1994). This recent origin is reflected in the linguistic and genetic homogeneity of Bantu (Excoffier et al., Supra, 1987; Weber et al., Supra, 2000). Thus, available historical, linguistic and genetic evidence suggests that only a subset of diversity found in sub-Saharan Africa contributed to the African American gene pool, and possible heterogeneity Sexual problems indicate that the diversity of all continents in Africa is much less than that present in modern African American populations. Unfortunately, the degree of heterogeneity that exists in West and Central Africa remains largely unknown due to the lack of available information about populations in this region.

  The degree of heterogeneity within the European population and within West and Central Africa remains largely unknown, so the potential impact of heterogeneity needs attention, especially when considering mixed mapping approaches There is. There are two levels where heterogeneity can affect mixed mapping attempts. First, the heterogeneity leads to an incorrect estimate of the parent's frequency for the markers used in the map, and biases the mixture estimate accordingly. If the goal of mixed mapping is to infer linkage conditioning in the parental mix, it is important to avoid misspecification of ancestor-specific allele frequencies because this is the final conclusion of the analysis. It can have an impact. Second, heterogeneity can affect the number of loci for the phenotype to be studied.

  The impact of heterogeneity in biasing estimates of genetic contribution to a mixed population can be reduced by selecting markers that show uniformity within the main parent population (European and African). sell. In this way, the problem of contribution to the parent population of different geographical regions is minimized, reducing the bias in the mixture estimation. This strategy has been carried out in previous mixed studies (Parra et al., Supra, 1998, 2001; Pfaff et al., Supra, 2001) and systematic markers that may provide information in different European and African populations. Was analyzed. As an example, information can now be provided in samples from five African populations, two from Nigeria, two from Sierra Leone, and one from the Central African Republic to test heterogeneity within Africa. Sexual markers were genotyped and markers that showed significant heterogeneity were excluded from the analysis (see below). All of these samples came from areas affected by slave trade. If desired, a sample from Angola, an area that was the source of about 40% of slaved Africans, may be incorporated, thus providing another sample of the African parent population . In addition to this strategy, it is important to note that there are statistical methods to test for misspecified parent frequencies (see, eg, McKeigue et al., Supra, 2000).

  With regard to possible problems of heterogeneity in the phenotype to be studied, heterogeneity due to the presence of synonymous genes (locus heterogeneity) that affect the phenotype can be any other mapping As is the case with the method, it is expected to reduce the power of mixed mapping to detect significant genotype effects. Heterogeneity can also be due to multifunctional alleles within a particular gene (allelic heterogeneity). One example is MC1R, where approximately six relatively common variants lead to red hair, freckles, and pale skin between native Europeans and their offspring populations. These variants within Europeans are in different haplotype backgrounds, thus reducing the ability to detect the effect of the MC1R gene in association analyzes compared to when a single mutation occurs and rises to high frequency Let However, in mixed populations (e.g. European / African), all of these variants are allelic with markers that provide information about ancestry (e.g., MC1R markers, see Table 1). Because all of them have the effect of lightening the skin, their information has identified six functional mutations in the identification of MC1R by mixed mapping rather than having only one functional mutant specific to Europeans It is compromised that the body is not different. As long as the effects of functional variants within a particular parent population are in the same direction (e.g. in reducing the risk of disease), allelic heterogeneity does not appear to be a serious problem in mixed mapping It is.

  The majority of genetic variation between human individuals (80-90%) is between individuals; only 10-20% of the variation is due to population differences (eg Nei, supra, 1987; Cavalli-Sforza et al., Supra, 1994; Deka et al., Supra, 1995). Most populations share alleles, and those alleles that are most frequent in one population are generally also frequent in others. Very few classical (blood group, serum proteins and immunological) or DNA genetic markers that are either population specific or have a large frequency difference between geographically and ethnically defined populations Less (Roychodhury and Nei, supra, 1988; Cavalli-Sforza et al., Supra, 1994). Despite this apparent lack of unique genetic markers, significant physical fitness among human populations, probably reflecting long-term adaptation to unique ecological conditions, accidental genetic drift and gender selection And there are physiological differences. In modern populations, these differences are evident both in morphological differences between ethnic groups, as well as differences in susceptibility and resistance to disease.

The unique alleles most useful for mixing and mapping studies are also those that have large differences in allele frequencies between populations (Reed, supra, 1973; Chakraborty et al., Supra, 1992; Stephens et al., Supra, 1994). . The fact that they are completely absent from all other populations can simplify some of the statistical calculations and facilitate more confident parental allele frequency estimations, Not the main reason. The name population-specific allele (PSA) was initially used to describe genetic markers with large allelic frequency differences between populations (Shriver et al., Supra, 1997; Parra et al., Supra, 1998). ), These markers are now called Ancestry Informative Markers (AIM) in more correct and descriptive terms. For biallelic markers, the frequency difference (δ) is equal to p _x -p _y (which is equal to q _y -q _x ), where p _x and p _y are the frequencies of one allele in populations X and Y , Q _x and q _y are the other frequencies. Median δ levels between major ethnic groups range between 15% and 20%, and the majority of arbitrarily identified biallelic genetic markers (> 95%) have δ <50% (Dean et al., Supra, 1994, incorporated herein by reference). Statistical estimates of power in mixed mapping studies based on using markers with Fst> 0.4 have been presented previously (McKeigue et al., Supra, 2000). With 1,000 such markers evenly spaced across the genome, 80% statistical power to identify disease genes that account for twice the relative risk between parental populations It was proved possible to have.

  AIM and their use by the methods of the invention are demonstrated in Examples 1-6 (below). In addition, allelic frequency data were reported for markers that inform mixed mapping in African American and Hispanic populations (Smith et al., Supra, 2001; Collins-Schramm et al., Supra, 2002). In order to apply the method of the present invention to the analysis of disease predisposition or drug responsiveness, estimation of mixing ratio and mixing kinetics is important. Controlling the genetic structure in a mixed population requires knowledge of the ancestral proportions and genetic structure of these populations. Reliable estimation of the mixing ratio may allow identification with information on the population to consider. Since mixed LD caused during hybridization depends on the level of mixing, sampling should generally be concentrated in those areas of the country where there was more mixing.

  In addition to knowing the proportion of ancestors, it is important to understand the level of population structure that exists in the population under consideration. A uniform population is an arbitrary mating population that is formed with no similar mating, with more or less families, random combinations, and regardless of DNA genotype. Uniformity is expected and found in most large groups of people from all over the world. However, if stratification exists in a population where individuals do not mate randomly, the population will not be uniform. Mixing is one possible mechanism for introducing genetic structure in a population, and taking this genetic structure into account facilitates mixed mapping.

  The effects of genetic structure can be considered at two levels. First, the parent population is evaluated to determine if they show heterogeneity in the allelic frequency of the selected AIM; heterogeneity is mixed as discussed above May affect percentage estimates. Some methods can detect the presence of genetic structure. These methods are categorized into two main categories, named genome control (GC) methods (Devlin and Roeder, supra, 1999) and structuring-related (SA) methods (Pritchard and Donelly, supra, 2001). sell. Both methods, as discussed above, estimate and correct for the effects of genetic structure that may have been due to sampling effects or to the real demographic layer in the sampled population Requires genotyping of a panel of unlinked markers to do. The SA method (Pritchard et al., Supra, 2000; Pritchard and Donelly, supra, 2001) was used to test for genetic structure in the parental population. This method was implemented with a software program available from Jonathan Pritchard based on using genotype information provided by unlinked markers to infer population structure. In addition, to test for the existence of the structure, the program estimated the ancestry percentage of individuals, and for this study this Bayesian method was used to complement the maximum likelihood estimation method. These two methods result in an estimate of the ancestry of individuals that are highly correlated.

  The second source of genetic structure in the mixed population is for the mixing process itself, in which newly induced linkage disequilibrium is introduced into the mixed set. AIM, such as that exemplified herein, is a sensitive indicator of population structure, particularly related to ancestral proportions. To assess the presence of population structure, samples were tested for non-random associations of alleles both within loci (Hardy-Weinberg disequilibrium) and between loci (gamete imbalances) to estimate individual ancestry (See Pfaff et al., Supra, 2001; Parra et al., Supra, 2001).

  The history of African Americans can be traced back to 1619, the year when the first Africans arrived in the British colony (Jamestown), but as early as 1526, the existence of African slaves became the United States. Reported on a Spanish expedition to deaf places (South Carolina, Georgia, Florida and New Mexico). Organized slavery began just shortly after, but only in the early 18th century, slave imports accounted for a proportion of the demand for workers cultivating tobacco, indigo and rice plantations in the southern colonies. The peak occurred in the 10 years from 1790 to 1800 and in the early years of the 19th century. In 1808, slave trade became illegal, but it continued at a low rate for decades. Different estimates have been given for the total number of slaves brought to the United States, with numbers generally in the range between 380,000 and 570,000.

  Although it is difficult to accurately determine the ethnic origin of African slaves, information from the ship list provided an approximate picture of their geographic origin. The slave trade affected a very large area of West and Central West Africa, primarily the coastline between Senegal in the northern part of today's country and Angola in the southern part. The most important areas are from Sene Gambia (Gambia and Senegal), Sierra Leone (Guinea and Sierra Leone), Windward Coast (Ivory Coast and Liberia), Gold Coast (Ghana), Benin Bay (Volta River) The Benin River), the Biafra Bay (from the east of the Benin River to Gabon), and Angola (southwest Africa including parts of Gabon, Congo and Angola). Curtin (supra, 1969) presents an estimate of proportional contribution by region based on data from the British trade in the 18th century (the peak of the Atlantic slave trade). It was shown to be the region giving the highest number (about 25% each). However, there were significant differences in ethnic origin depending on the port of entry in the United States, and figures for the Virginia and South Carolina colonies differed considerably.

  African American history is characterized not only by forced migration from Africa, but also by a mix with other ethnic groups that they met when they arrived in North America, including Europeans and Native Americans. It was. However, there are few historical records dealing with the problem of mixing. In addition, during the period from the abolition of slavery to the present, there were important factors that formed the current African-American population. Of special interest is the pattern of African American migration 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 great migration from the southern part of the countryside to the northern urban areas that began after World War I, are particularly relevant. It had enormous influence in defining the current distribution of the American population (Johnson and Campbell, Black Migration in American: A Social Demographic History; Duke University Press, Durham, NC, 1981).

  With respect to Hispanic, the term “Hispanic” is created primarily for political demographic purposes and is generally used to identify people of Latin American origin or ancestry living in the United States. . Although this definition brings together people with very different historical, cultural and linguistic backgrounds, this classification has been widely used. Central America, the Caribbean and South America have been under the control of the Iberian regime (Spain and Portugal) for centuries, but they have completely different regional histories before and after the colonial era. It was. Populations from four continents, North and South America, Europe and Africa, contributed to the formation of modern Hispanic populations. Here are the anthropological backgrounds of the three major Hispanics currently living in the United States-Mexican-Americans, Puerto Ricans and Cuban-Americans, together making up more than 80% of the total Hispanic population Be considered.

  Mexican Americans show the highest American Indian contribution of the three aforementioned groups. Shortly after Spain's conquest of Mexico at the beginning of the 16th century, the mixing of Spanish men with American Indians resulted in a mixed population (Mestiso) of increasing importance, this racial mix Continuing through the third century of Spanish rule in the New Spain, a Mexican population was formed both biologically and culturally. The majority of estimates showed American Indian components in Mexican Americans ranging between 30% and 40% (Hanis et al., Supra, 1986; Long et al., 1991; Hanis et al., Diabetes Care 14: 618 -627, 1991; Merriwether et al., Amer. J. Phys. Anthrop. 102: 153-159, 1997). In addition, it is interesting to point out that several studies have shown differences in the amount of American Indian ancestry depending on socio-economic status (Chakraborty et al., Genet. Epidemiol. 3: 435-454, 1986; Mitchell et al., Ethnicity and Disease 3: 22-31, 1992). There were also substantial Africans in the Mexican territory during Spanish rule. Curtin (supra, 1969) estimated that the total number of Africans imported into Mexico during the entire period of slave trade was approximately 200,000. However, their contribution to the Mexican gene pool was estimated to be much lower than that of Europeans and American Indians, ranging from zero to 10% (e.g., Hanis et al., Supra, 1991. reference).

  In the Caribbean colonies (Cuba and Puerto Rico), the situation was very different from the mainland. Indigenous American populations were much smaller and soon died of suffering and illness soon after the first contact with Europeans. Nevertheless, the proportion of mixing during the early period of colonization resulted in an appreciable genetic contribution (about 18%) from Arawak and the Caribbean, natives of the Hispanic Caribbean. High enough (Hanis et al., Supra, 1991). Another differential feature in this area is the significant African influence, which is also reflected in many aspects of the country's current society such as Cuba, Puerto Rico and the Dominican Republic. African slaves have been imported to work on sugar plantations and have even surpassed European populations (Kanellos and Perez, Chronology of Hispanic-American history: from pre-Cloumbian times to the present New York, Gale Research, 1995). Thus, the percentage of African genetic contribution in modern Cubans (20%) and Puerto Ricans (37%) is significantly higher than in other Hispanic populations (Hanis et al., Supra, 1991).

  Race is a complex concept that reflects both cultural and biological characteristics of a person or group of people in common usage. Given the fact that physical differences between groups are often accompanied by cultural differences, it was difficult to separate these two elements. There were moves in some areas of science that state that race was just a social construct, and that the issue was oversimplified. This can often be true, depending on what aspect of variability among people is to be considered, but is false for many specific cases of differences between populations in the world sell. One obvious example of a biological difference is skin color. Culture or environment has little effect on the level of pigmentation in the human skin. There are still dramatic differences across the group. Pigmentation is, of course, just the skin and is quite simple in terms of the complex environment in which we live and how they affect the quality of life of individuals and groups.

  Human species are relatively young, most likely to originate in East Africa 100,000 years ago as species, diverging as groups and settled on Earth (Cavalli-Sforza and Cavalli-Sforza, The Great Human Diasporas. History of Diversity and Evolution (Perseus Books, Cambridge, MA, 1995)). During these migrations and since then, there have been some independent developments of groups settled on various continents of the world. The simplest evidence of this progress is in the differences in allele frequency at genetic markers. In general, alleles found in one population are also found in all populations, and the most common alleles in one population are also common in others. These similarities between populations shed light on the recent common origin of all populations. However, there are examples of genetic markers that differ between populations, and as disclosed herein, these markers, AIM, can be used to estimate the origin of a human or population ancestor.

  The present invention provides a method for estimating the proportional ancestry of at least two ancestry groups of a test individual, and in particular, provides a confidence level for the proportional ancestry. The method of the present invention allows a sample comprising a nucleic acid molecule of a test individual to hybridize oligonucleotides capable of detecting the nucleotide occurrence of SNPs in a panel of at least about 10 AIMs indicating a BGA for each ancestral group being examined. An ancestor to be tested with a predetermined confidence level, wherein the contacting step is under conditions suitable for detecting the nucleotide appearance of the AIM nucleotide of the test individual by the hybridizing oligonucleotide; It may be performed by identifying a population structure that correlates with the nucleotide occurrence of each AIM of the group, wherein the population structure indicates a proportional ancestry.

  The term “biogeographical ancestry” or “BGA” is used herein to describe the biological or genetic components of a race. BGA is a simple and objective description of the origin of a person's ancestry in the language of the main population groups (eg, Native Americans, East Asians, Indo-Europeans, and Sub-Saharan Africans). BGA estimates can represent the mixed nature of many people and groups today. In many countries, including the United States, there was extensive mixing among initially segregated populations. The term “admixture” is used herein to refer to such population mixing. In this regard, BGA estimation can be understood as a mixture of individuals, taking the form of a series of percentages that add up to 100%. For example, a person may have 75% IndoEuropean, 15% African and 10% Native Americans, or 100% IndoEuropean ancestry.

  Proportional ancestry estimated by the method of the present invention is the proportion of any ancestry group including, for example, sub-Saharan African, Native American, Indo-European, East Asian, Middle Eastern, or Pacific Islander ancestry groups It is generally a combination of two or more such ancestry groups. Thus, the proportional ancestry of a test individual may include a proportional affiliation between sub-Saharan African and Indo-European ancestry groups (eg, 80% sub-Saharan and 20% Indo-European; or 60% Saharan). South Africans, 20% Indo-Europeans and 20% third ancestry groups); Indigenous Americans and Indo-European ancestry groups; East Asian and Indigenous American ancestry groups; Indo-European and East Asian ancestry groups And so on.

  A panel of AIMs useful for estimating an individual's proportional ancestry is the AIM shown in SEQ ID NOs: 1-331, for example, proportional including Indo-European, Sub-Saharan African, East Asian and Native Americans AIM set forth in SEQ ID NOs: 1-71 which may be useful for measuring ancestry; or SEQ ID NOs: 7, 21, which may be useful for measuring proportional ancestry of East Asians and sub-Saharan Africans 23, 27, 45, 54, 59, 63 and 72-152; or SEQ ID NOs: 3, 8, 9, 11, 12 which may be useful for measuring the proportional ancestry of East Asians and Indo-Europeans 33, 40, 59, 63 and 153-239; or SEQ ID NOs: 1, 8, 11, 21, 24, 40, which may be useful for determining the proportional ancestry of Indo-Europeans and Sub-Saharan Africans 172 and 240-331, AIM.

  The estimation is made, for example, on the individual's proportional ancestry for the three ancestry groups. In this method, identifying the population structure within an individual that correlates with the AIM nucleotide appearance of the tested individual is the ability of sub-Saharan African ancestry, Native American ancestry group, IndoEuropean ancestry group and East Asian ancestry group. Determining the likelihood of membership for each; then selecting the three ancestry groups with the largest likelihood value for the individual; all possible proportions among the three ancestry groups with the largest likelihood value Determining the likelihood of global affiliation, thereby identifying a population structure or proportional affiliation that correlates with the AIM nucleotide appearance of the test individual; and only one proportional combination of maximum likelihood Can be carried out by identifying. Alternatively, identifying a population structure that correlates with the nucleotide occurrence of AIM involves performing six binary (binary) comparisons, including a likelihood determination for each group affiliation compared to each other group; Then select the three ancestry groups with the largest likelihood values across all comparisons; determine the likelihood of all possible proportional affiliations among the three ancestry groups with the largest likelihood values Carried out by identifying a population structure or proportional affiliation that correlates with the AIM nucleotide occurrence of the test individual; and identifying only one proportional combination of maximum likelihood sell. Such a method works for individuals who are 100% belonging to a single group in addition to ternary individuals.

  Estimating the proportional ancestry of an individual, including the proportions of the three ancestry groups, also performs three ternary comparisons between groups; all possible proportions among the three ancestry groups with the highest likelihood values Determining the likelihood of global affiliation, thereby identifying a population structure or proportional affiliation that correlates with the AIM nucleotide appearance of the test individual; and a single proportional combination of maximum likelihood This can be done by identifying. The advantage of this method is that a graphical representation of a comparison of three ancestor groups can be created, the graphical representation comprising triangles, each ancestor group being represented independently by a triangle vertex, and for individuals The maximum likelihood value of the proportional affiliation of includes the points in the triangle (see FIGS. 2 and 3). If desired, the graphical representation may further include confidence contours that indicate the confidence level associated with estimating proportional ancestry.

  An estimate of an individual's proportional ancestry can also be made where the proportional ancestry includes a proportion of four ancestry groups. In various aspects of the method, identifying a population structure that correlates with the AIM nucleotide occurrence of the test individual comprises performing six binary comparisons, or three ternary comparisons between groups, Or performing one quaternary comparison; determining the likelihood of all possible proportional affiliations among the four ancestry groups with the largest likelihood values, thereby determining the AIM of the test individual A population structure or proportional affiliation that correlates with nucleotide occurrence is identified; and identifying only one proportional combination of maximum likelihood. If desired, the method may further comprise creating a graphical representation of the comparison of the four ancestry groups, the graphical representation comprising a pyramid, each ancestor group being independently represented by a pyramid vertex; And the maximum likelihood value of proportional affiliation for an individual includes a point in the pyramid. If desired, the graphical representation may further include a confidence contour that includes a sphere centered at that point, where the sphere indicates the confidence level associated with estimating the proportional ancestry.

  As disclosed herein, such methods are useful, for example, as a forensic tool. Using DNA samples obtained at crime scenes, the method can provide investigators with expected information about the likelihood of individual ancestry in addition to hair, skin and eye pigmentation, so the method Provides substantially more information to forensic medicine. In comparison, current DNA methods require retrospective information because they require that DNA samples from crime scenes be included in the database or compared with DNA samples taken from specific individuals. Only give. Thus, the latter method can provide confirmation that the suspect is likely to be a criminal, but the suspect is not Do not provide useful information until arrested.

  As disclosed herein, a method for estimating the proportional ancestry of a test individual is also a tool that can supplement phylogenetic information that is generally based on relationships established using geopolitical information. (See Example 3). For example, the method provides information that can be used to create a global ancestry map, and the location of the population with proportional ancestry corresponding to the proportional ancestry of the test individual is indicated in the ancestor map. As such, the method involves overlaying an ancestor map with a phylogenetic map, where the pedigree map indicates the location of a population that has geopolitical relevance with respect to the examined individual, and the most likely family of the examined individual. The method may further include statistically combining the information of the ancestor map and the phylogenetic map so as to obtain a high estimation.

  By the method of the present invention, identifying a population structure that correlates with the nucleotide occurrence of AIM can be achieved by comparing the AIM nucleotide occurrence of the tested individual with a known proportional ancestor corresponding to the nucleotide occurrence of AIM that indicates BGA. Can be done. Known proportional ancestry corresponding to the nucleotide occurrence of AIM indicating BGA may be included in the table or other list, and the nucleotide occurrence of the test individual may be visually compared to the table or list, or included in the database In other words, the comparison can be made electronically, for example, using a computer. A particularly useful application of the method of the present invention is to combine a known proportional ancestry corresponding to the nucleotide occurrence of an AIM indicative of BGA with a photograph of the person from whom the known proportional ancestry has been determined, and accordingly the body of the examined individual Providing a means for further inferring the characteristic. In one aspect, the photograph is a digital photograph and includes digital information that may be included in a database that may further include a plurality of such digital information of the digital photograph, each of which represents an AIM nucleotide occurrence that indicates a person's BGA Associated with a known proportional ancestor.

  The method of the present invention may further comprise identifying a photograph of a person having a proportional ancestry that corresponds to the proportional ancestry of the test individual. Such identification can be done by manually examining one or more files of the photograph, which are organized according to, for example, the nucleotide occurrence of a person's AIM in the photograph. Identifying the photos also includes scanning a database containing multiple files, each file containing digital information corresponding to a digital photo of a person with a known proportional ancestor, and the BGA of the individual being examined. This may be done by identifying at least one photograph of a person with an AIM nucleotide occurrence showing a BGA that matches the indicated AIM nucleotide appearance.

In accordance with the present invention, BGA can be measured using any of several variations of the disclosed BGA test, and a selection of ancestral informative markers (AIMs) characterized in a number of well-defined population samples. This includes three BGA tests called ANCESTRYbyDNA ™ 1.0 test, ANCESTRYbyDNA ™ 2.0 test, and ANCESTRYbyDNA ™ 3.0 test (DNAPrint genomics, Inc .; Sarasota, FL). AIM is selected on the basis of displaying substantial differences in frequency between population groups, thus providing information about the origin of a particular person whose ancestors are otherwise unknown. For example, the Duffy Null allele (FY ^* 0) is very common in all sub-Saharan African populations (approximately equal to colonization or 100% allele frequency) but is not found outside Africa. Thus, a person with this allele is very likely to have some level of African ancestry. In the analysis of AIM in DNA samples from people of unknown origin, the likelihood (or probability) that a person is from a particular parent population can be determined by calculating all possible mixtures of the parent population . The population (or population combination) with the highest likelihood is taken as the best estimate of the person's ancestry percentage; confidence intervals in these point estimates of the ancestry percentage are also calculated.

  Objective assessment of the biological components of human ancestry provides important knowledge about the person whose DNA is being examined. For example, analysis of ancestral biological components can identify genetic contributions to, for example, higher rates of hypertension and diabetes in African Americans, or higher rates of dementia in European Americans Can elucidate health disparity. BGA estimates can also help individuals separated by adoption or some other incident to meet their ancestry population. Even if a person is not particularly moved to meet an ancestor again, he or she may uncover their family's past, for example, verify family legends or identify forgotten roots. it can. Because the disclosed method is based on DNA analysis, it provides a personal demographic tool that, unlike census, can provide highly accurate demographic data.

  There are several commercially available tests that analyze mitochondrial DNA (mtDNA) or Y chromosome markers that have been promoted as a means of knowing the origin of one ancestor. Although these tests can provide information about the origin of some of the ancestors of a person, 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; whereas ten generations ago, a person has 1024 Has ancestry. The 10th generation is roughly 250 years long, especially within the time frame that is systematically targeted, especially when considering the settlement of North America, for example. Because mtDNA and Y chromosome tests only look at small parts of the genome (maternal and paternal strains, respectively), they can only provide information about a very small part of a person's ancestry. The BGA test of the present invention utilizes sequences throughout the human genome and can therefore provide information about a greater number of ancestors.

  Accordingly, the present invention provides a method for estimating a proportional ancestor of at least two ancestry groups of a test individual with a predetermined confidence level. Such a method is referred to as a “biogeographical ancestry test” or “BGA test”, but for example, a sample containing nucleic acid molecules of a test individual is at least approximately indicative of a BGA for each ancestral group being examined. Contacting a hybridizing oligonucleotide capable of detecting the nucleotide occurrence of a SNP in a panel of 10 AIMs, wherein the contacting step detects the nucleotide occurrence of AIM in the test individual by the hybridizing oligonucleotide. Identifying a population structure that correlates with the nucleotide occurrence of each AIM in the ancestral group being examined, with a predetermined confidence level, wherein the population structure identifies a proportional ancestor The steps shown can be performed.

  As used herein, the term “proportional ancestry” refers to the percent contribution of each ancestry group (if more than one) to which an individual belongs. Proportional ancestry estimated by the methods of the present invention can be any ancestry group including, for example, the proportion of sub-Saharan Africans, Native Americans, Indo-Europeans, East Asians, Middle Easterns, or Pacific Islander ancestry groups. It can be a proportion, generally a combination of two or more such ancestry groups. Thus, the proportional ancestry of a test individual may include the proportion of sub-Saharan African and Indo-European ancestry groups (eg, 80% sub-Saharan and 20% Indo-European; or 60% sub-Saharan Africans). , 20% Indo-European, and 20% third ancestry groups); or Native American and Indo-European ancestry groups; East Asian and Indigenous American ancestry groups; Indo-European and East Asian ancestry groups, etc. May include percentage. Similarly, proportional ancestry includes Native American, East Asian and Indo-European ancestry groups; Sub-Saharan African, Native American and Indo-European ancestry groups; Sub-Saharan African, Native American and East Asian ancestry It can include proportions such as groups.

  A panel of AIMs useful for estimating an individual's proportional ancestry is the AIM shown in SEQ ID NOs: 1-331, for example, proportional including Indo-European, Sub-Saharan African, East Asian, and Native Americans An AIM set forth in SEQ ID NOs: 1-71 that may be useful for measuring ancestry. For example, the AIMs shown in SEQ ID NOs: 7, 21, 23, 27, 45, 54, 59, 63 and 72-152 are useful for measuring the proportional ancestry of East Asians and sub-Saharan Africans The AIMs shown in SEQ ID NOs: 3, 8, 9, 11, 12, 33, 40, 59, 63 and 153-139 are useful for measuring the proportional ancestry of East Asians and Indo-Europeans. And the AIMs shown in SEQ ID NOs: 1, 8, 11, 21, 24, 40, 172 and 240-331 are useful for measuring the proportional ancestry of IndoEuropeans and sub-Saharan Africans sell.

  The ANCESTRYbyDNA ™ 1.0 test (DNAPrint genomics, Inc.) is the first version of the BGA test specifically designed to provide information on the percentage of ancestry at the continental level. As such, the ANCESTRYbyDNA ™ 1.0 test allows information to be obtained regarding the levels of indigenous American, European and African ancestry as three component groups. The ANCESTRYbyDNA (TM) 2.0 test compared, indigenous Americans, Indo-Europeans (including South Asians such as Europeans, Middle Easts and Indians), Africans and East Asians (including Pacific Islanders) ) For most continents, including percentages of ancestry at the continental level, and can distinguish ancestry within Asia and the Pacific Rim. The ANCESTRYbyDNA (TM) 3.0 test can further limit the level of ancestry within a continent, for example, by distinguishing between Japanese and Chinese, or Northern Europeans and the Middle East, and accordingly in certain continents It gives a deeper insight into where human ancestry comes from.

  The ANCESTRYbyDNA (TM) 2.0 test has been logically divided into four BGA diagrams, with South Asians, Middle Easts and Europeans being classified into a single group called Indo-Europeans (see Example 2). ). This classification is anthropological evidence and cultural ties between these groups (eg, their language comes from a common maternal body). The results disclosed herein demonstrate that these groups are much more similar to each other in gene sequence content than the other groups. ANCESTRYbyDNA (TM) 2.0 testing is also performed more accurately when Pacific Islanders are classified with East Asians. As such, the four classifications used in the ANCESTRYbyDNA ™ 2.0 test are: 1) Native Americans (ie, those who moved and resided in South and North America); 2) Indo-Europeans (European, Middle Eastern and South Asians like Indians); 3) East Asians (Japanese, Chinese, Korean, Pacific Islander); and 4) Africans (sub-Saharan). The ANCESTRYbyDNA (TM) 3.0 test can further distinguish between South Asians and Europeans, and Pacific Islanders and East Asians, and accordingly 6 parts (Indigenous Americans, Europeans, Africans, South Asians) Human, East Asian and Pacific Islander), but the confidence interval is larger than that obtained with the ANCESTRYbyDNA ™ 2.0 test. Further improvements to testing are given and confidence intervals are decreasing. By analyzing the supplemental panel and thereby improving the confidence interval by about 50%, the confidence interval centered around the point estimate can be reduced and the accuracy of the test can be increased accordingly.

  The algorithm used to determine the proportion of ancestors was developed based on the idea that certain statistical methods can be used to infer the ancestry proportions in an individual's sample based on their sequence. (See Example 6; see also Table 12). The method of making this inference using this algorithm is similar to that of others, and if the allele frequency in a population is known and this frequency is significantly different from population to population, then "maximum likelihood estimation" The “value” (MLE) can be used to determine the probability that a person with that allele belongs to one of the groups. The process is the same when expanded to include multiple alleles from multiple loci and multiple populations. As a simplification, Bayes' theorem is that the probability of an event assuming a situation (called the posterior probability) is a function of the frequency of the situation assuming the event (conditional probability) and the frequency of the event itself (prior probability). It is said that. By determining the probability of an event assuming a situation for a wide range of possible events, the one with the highest probability is selected and the MLE for the probability can be obtained accordingly.

  In this algorithm, events are proportional to ancestors and situations are individual genotypes. If the minor allele frequencies for 10 SNPs in two populations of humans are known and the human sequence in each of the 10 SNPs is known, a simple to one of the two groups Binary classification can be done by selecting the one with the highest conditional probability. This represents little improvement over current methods for measuring BGA from DNA samples. What is provided by the present invention is the ability to obtain ancestral proportions for more complex and realistic scenarios of ancestors. Many like 99% African, 1% European, 0% Native American, 0% East Asian; or 98% African, 1% European, 1% Native American, 0% East Asian, etc. There are possible combinations. The posterior probabilities for each of thousands of possibilities are not the same for any particular individual, given his or her multilocus genotype (i.e., many AIM genotypes), in fact, each genotype Have the highest posterior probability or likelihood. The algorithm chooses this combination (ie, MLE).

  The previous method was limited in that it did not know the confidence of the estimate. The algorithm addresses this limitation by plotting MLE graphically, including plotting a confidence region centered on MLE so that the confidence level can be ascertained (see Figures 2 and 3). . Furthermore, algorithms that perform MLE computations (ie, software code) operate in a significantly efficient manner. The triangle plot provided by the algorithm is an ingenious way to graph MLE calculations and their confidence intervals. To read the triangle plot (see below), a perpendicular is dropped from each vertex of the triangle (triangle point) to the opposite side (bottom) of the triangle (see FIG. 2A). In this figure, the circle represents MLE, and the line is down from the Native American (NAM) vertex to the bottom line; the line is indigenous, from 0% at the base to 100% at the vertex (or tip). Serves as a measure of the percentage of American ancestry. Projecting a circle onto this line can be inferred to hold the flashlight to the right side of the triangle at the same level as the circle and to observe the shadow that 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 approximately 15% Native American, as indicated by the hash mark on the line.

  The results given using the disclosed method are statistical estimates of the BGA mixture for individuals (maximum likelihood estimation), shown as points on a triangle plot representing the proportion of the three groups most relevant for the individual. Value (MLE)). While MLE is the most likely estimate, the true value for an individual can be a different set of proportions. Triangular plots with calculated and plotted estimates that are two, five and ten times less likely than MLE are illustrated. The first contours centered on the MLE define the extent of the space where the estimate is likely to be doubled, locating those locations near the line reflecting almost the value of 2 and those near the MLE of almost 1. Contain; second contour centered around MLE delimits the space where the estimate is five times less likely in the same stepwise manner, continuing from the first contour to the second contour. The third contour line defines the range of the space where the estimated value is less likely from 5 times (near the second contour line) to 10 times (near the third contour line). The more you read the number of DNA positions, the closer these contours will be to the MLE point. On a triangular plot, the likelihood (probability) that a true value is represented at a point different from MLE increases until it hits MLE, where the probability is the highest (ie, maximum likelihood estimate; MLE) . The assay can be performed so that the contour lines are very close to the MLE by sequencing a very large set of markers. However, in order to keep the assay affordable and efficient, the survey will show that the desired number of markers (e.g. 10, 15, 20) is sufficient to measure the most likely proportion with good confidence. , 25, 30, 40, 50, 60, 70, 80, 90, 100 or more). In this regard, a variety of different panels of 100 SNP markers have been examined, 71 AIM panels are used in numerous studies, and 175 AIM panels achieve true confidence It is supposed to be examined as follows.

  The BGA test of the present invention has been validated by measuring the frequency of DNA sequence variants in various human populations. Furthermore, the tests were evaluated using a large number of people from a wide range of ancestry groups, and the estimates were in good agreement with those known from anthropological and historical data. For example, Hispanics are known to have been born as ethnic groups from a mix of colonial European and Native Americans, and hundreds of Hispanics examined using BGA testing are almost exclusively these The two groups were aligned. As another example, Nigerians plot as having almost pure African BGA, but African Americans are expected from knowledge of the mix between Africans and Europeans in the United States But plot more as a mix between this group and Europeans.

  The method was also validated through a pedigree challenge (see Example 1); that is, if BGA is determined from the mother and father, those of those children should plot somewhere between the two. A number of pedigrees were examined using the test, and the percentage of children's ancestry was always plotted between those of the children's parents. If MLE estimates are examined objectively (blindly), they prove to be excellent estimates of ancestry proportions. For example, data for a European-American male whose mother is a mixed-European mother and whose father is mostly Greek shows that the male has 85% European ancestry but also 15% Native American ancestry. (Example 1). In fact, his paternal great-grandmother was a purebred Cherokee, thus confirming the results of the test (based on the laws of genetics, the man was 100% native American with his great-grandmother, and If none of his other relatives had indigenous American ancestry, it is expected to have 12% indigenous American ancestry). In addition, the male wife was Mexican and she was measured to be mostly indigenous Americans, but had some indigenous American and African heritage. This was also expected based on what is known from Hispanic anthropological origins derived from the marriage of Spanish explorers and Native Americans in the colonial Caribbean and Latin America. Each of the three male and female children, as expected, plotted approximately at the midpoint between the parents. None of the children showed any Asian or Pacific Islander ancestry, but neither would have been possible because neither of the parents showed any significant Asian or Pacific Islander heritage (assuming accurate testing). ), It was found that none of the children had more African ancestry than their mothers, but that would also not be possible given the fact that the father has virtually no . Thus, the child's results were consistent with those of the parents, and the MLE value was an accurate estimate when tested against what is known from the biographical data.

  So far the genotype (nucleotide character) measured is quite accurate. Since the latest available gene readers are used, accuracy greater than 99% accuracy is routinely achieved for each site. If an exact value is not obtained for a particular site in a particular sample, “FL” is shown instead of the genotype letter for that site. Having a little FL generally does not interfere with good ancestry estimation. A sample can produce FL for a site because, for example, a small region of the chromosome around this site is missing or has different sequence characteristics than most; or a DNA sample is collected This is because sufficient DNA could not be obtained from the buccal application used to do this.

  The genome was scanned for a useful panel of BGA AIMs, and the best 71 AIMs identified using the maximum likelihood algorithm to measure the BGA mixing ratio were selected (Table 6). Using these AIMs, the majority of BGA affiliations can be measured in a manner consistent with the self-contained concept of BGA, and the BGA mix ratio is significantly improved compared to previously described methods for racial inference Was able to be measured with the accuracy, accuracy and reliability provided (see Example 2; see also Example 1 with 32 marker tests). This test can be used during research design to help reduce or eliminate the devastating effects imposed by hidden or small population structures. Testing can also be useful for forensic scientists who currently use inaccurate and sometimes inaccurate means to infer race from crime scene DNA.

  The present invention also provides a product comprising one or more photographs, each photograph being of a person with a known proportional ancestry corresponding to a population structure containing the nucleotide occurrence of AIM representing BGA. Proportional ancestry is associated with the photographs in the goods. Products of the invention (i.e., photos and proportional ancestry information) can be contained in one or more files (e.g., photos and information in one file, or photos in one file, and links to photos) Information to a second file that may or may not be linked). If desired, more than one photo of an individual with a known proportional ancestry can be included in the same or linked file, eg, photos containing different profiles of individuals or photos of individuals at various ages.

  Similarly, multiple items (i.e., photos and proportional ancestry information) can be included in a single file, e.g., a file that includes multiple photos of different people, some or all of which have a BGA It has the same or different known proportional ancestry corresponding to the population structure containing the nucleotide occurrences of AIM shown. Such multiple items can also be included in different files, for example, each containing information about one photograph and the known proportional ancestry of the individual in the photograph, or two of each Contains more photographs, each of which contains the same known proportional ancestry, or each contains two or more photographs of different individuals, and some or all of the individuals Contains multiple files with different proportional ancestry compared to another individual contained in the file. Thus, multiple such items are provided, and multiple files are also provided, but each file has the same or different known proportional ancestry corresponding to the population structure containing the AIM nucleotide occurrences representing BGA Can include one or more items, i.e. photographs, which can be of one or more persons; and the plurality of files can each include one or more photographs of one or more persons And including one or more photographs of two or more different people, the different people may contain files that may have the same or different known proportional ancestry.

  A photograph of a person with a known proportional ancestry that corresponds to a product structure, ie, a population structure containing nucleotide occurrences of AIM representing BGA, can be a digital photograph, for a photographic image, and can be relevant or desirable Digital information including any other information (e.g., subject's answer to a questionnaire about the subject's age, name or contact information in the photo, or what the subject thinks about his or her ancestry) including. Such digital information of one or more digital photographs may be included in a database, and accordingly facilitate retrieval of photographs and / or known proportional ancestor information using electronic means. As such, the present invention further provides a plurality of products that each include at least two digital photographs each containing digital information. Where digital information about one or more items is included in the database, it includes, for example, computer hardware or software, magnetic tape, or computer disks such as floppy disks, CDs or DVDs, etc. It can be included on any medium suitable for containing a simple database. As such, the database can be contained by the computer, can contain the database, can accept the medium containing the database, or can access the database through a wired or wireless network, such as an intranet or the Internet. Can be accessed.

  The invention also provides kits useful for performing the methods of the invention. Such a kit may comprise, for example, a plurality of hybridizing oligonucleotides, each of at least 15 contiguous nucleotides of the polynucleotides shown in SEQ ID NOs: 1-331 (or polynucleotides complementary thereto) A plurality of which comprises at least 5 of such oligonucleotides (eg, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, etc.), each of which is SEQ ID NO: Based on different polynucleotides shown in 1-331. In one embodiment, the hybridizing oligonucleotide comprises at least 5 of the polynucleotide set forth in SEQ ID NOs: 1-71, or the polynucleotide complementary to any of SEQ ID NOs: 1-71. Contains consecutive nucleotides. In another embodiment, the hybridizing oligonucleotide is specific for at least 10 AIMs shown in SEQ ID NOs: 1-71. The kit of the present invention also includes at least five (for example, 5, 6, 7) specific to AIM shown in, for example, SEQ ID NOs: 7, 21, 23, 27, 45, 54, 59, 63 and 72-152. , 8, 9, 10, 11, 12, 13, 14, 15, etc.); or SEQ ID NOs: 3, 8, 9, 11, 12, 33, 40, 59, 63 and 153 A panel of at least 5 hybridizing oligonucleotides specific for AIM as shown in ~ 239; or specific for AIM as shown in SEQ ID NOs: 1, 8, 11, 21, 24, 40, 172 and 240-331 A panel of at least 5 hybridizing oligonucleotides; or two or more of such panels and / or of at least 5 hybridizing oligonucleotides specific for AIM shown in SEQ ID NOs: 1-71 Such hybridizing orifices, including panels It may contain at least two panels of gogonucleotides.

  The hybridizing oligonucleotides of the kits of the invention can include probes that are useful for detecting specific AIMs that include specific nucleotide occurrences at the SNP position of AIM; primers useful for primer extension reactions and nucleic acid amplification reactions Can include primers, useful primer pairs; or can include combinations of such probes and primers. The plurality of hybridizing oligonucleotides is not necessary, but is a nucleotide position of SNP or DIP of AIM, for example, nucleotide position 50 of AIM shown in any of SEQ ID NOs: 1 to 55 and 57 to 331, or a sequence Number: 56 of nucleotide 56, or for a complementary nucleotide sequence, may include a nucleotide corresponding to, such a hybridizing oligonucleotide may or may not be present at a particular nucleotide occurrence at the SNP position of AIM It is useful as a probe for identifying.

  The kits of the invention can also include at least one pair of hybridizing oligonucleotides useful for detecting the occurrence of nucleotides at the SNP position or the presence or absence of nucleotide sequences at the DIP position of the AIM. For example, a pair of hybridizing oligonucleotides comprises one oligonucleotide that hybridizes adjacent and upstream to the SNP position of AIM and a second oligonucleotide that hybridizes adjacent and downstream of the SNP position of AIM. In other words, one or the other of the pair further comprises a nucleotide complementary to the nucleotide occurrence suspected of being at the SNP position of AIM (i.e., one of the polymorphic nucleotides), and such a pair of hybrids These oligonucleotides are useful in oligonucleotide ligation assays. In another example, a pair of hybridizing oligonucleotides can include an amplification primer pair that includes a forward primer and a reverse primer, and such a pair of hybridizing oligonucleotides contains the SNP or DIP position of AIM. Useful for amplifying the polynucleotide portion comprising.

  The kit of the present invention may further comprise additional reagents useful for performing the method of the present invention. As such, the kit includes, for example, one or more polynucleotides comprising AIM, including a polynucleotide comprising AIM that is designed to be detected by a hybridizing oligonucleotide or a pair of hybridizing oligonucleotides of the kit. And such polynucleotides are useful as controls. Further, the hybridizing oligonucleotides of the kit can be detectably labeled, or the kit comprises one of the kits or different detectable labels that can be used to label the hybridizing oligonucleotide differently. Reagents useful for detectably labeling a plurality of hybridizing oligonucleotides can be included; such kits are for linking a label to a hybridizing oligonucleotide or for detecting a labeled oligonucleotide. And the like. The kits of the invention also include ligases, for example, when the hybridizing oligonucleotide of the kit includes a primer or amplification primer pair, a polymerase; or when the kit includes a hybridizing oligonucleotide useful for oligonucleotide ligation assays. , May be included. In addition, the kit can include appropriate buffers, deoxyribonucleotide triphosphates, and the like, depending, for example, on the specific hybridizing oligonucleotides included in the kit and the purpose for which the kit is to be supplied.

  The following examples illustrate but do not limit the invention.

Example 1
Measuring biogeographical ancestry using ancestral informative markers This example allows 32 ancestral informative markers (AIM) to estimate genetic contributions from African, European and Native American populations Demonstrate that

  The AIM used in the illustrated study includes single nucleotide polymorphism (SNP), deletion / insertion polymorphism (DIP) and Alu sequences (see Example 2 for identification of AIM). Markers were selected that showed differences between parental populations greater than 30% (Table 1; see also SEQ ID NOs: 332-363). Informative genetic markers can be used to confirm the utility of markers for mixed estimation, including Europeans (Spanish and German), Africans (from Nigeria, Sierra Leone, and Central African Republic) and Native Americans (Maya Identified by testing each candidate marker in a panel of populations (Spanish and Southwestern Native Americans).

マーカー名および染色体バンド、メガベース(Mb)での染色体上におけるマーカーのおおよその位置、ならびにアフリカ人とヨーロッパ人集団(AF/EU)、アフリカ人と先住アメリカ人(AF/NA)およびヨーロッパ人と先住アメリカ人(EU/NA)の間での頻度における差が示されている。30%より大きい差は太字で示されている(参照として本明細書に組み入れられている、Shriverら、前記、2003も参照)。
^＊ 括弧内の数字はAIMについての配列番号である；NS - 示されていない配列。 (Table 1) Ancestral information provision marker panel

Marker name and chromosome band, approximate location of the marker on the chromosome in megabase (Mb), and African and European population (AF / EU), African and Native American (AF / NA) and European and Indigenous Differences in frequency between Americans (EU / NA) are shown. Differences greater than 30% are shown in bold (see also Shriver et al., Supra, 2003, incorporated herein by reference).
^* Numbers in parentheses are SEQ ID NOs for AIM; NS-sequences not shown.

  Publicly available human genome sequence databases and polymorphism databases were screened to identify SNPs that meet the criteria for being a good AIM. Allele frequencies are available for many of its SNPs in public databases for three populations-African, European and Asian. Since these frequencies are derived from a small number of samples, they are not always accurate. The primary criterion for selection herein is the delta value derived from using these frequencies and is a statistical measure of the difference in minor allele frequencies between various populations of humans. For example, a C or G polymorphism at a particular location in the human genome has a high delta value, although C is predominantly present in individuals of European descent and G is predominantly present in individuals of Native American descent. Therefore, it has a good AIM qualification. Similarly, the A or C polymorphism at a particular location in the human genome is large between these groups, although A is predominantly present in individuals of African descent and C is predominantly present in individuals of Asian descent. It has a frequency difference and hence a high delta value and is therefore a good AIM qualification. A list of such “candidate” AIMs is collected and ordered from the largest delta value to the smallest delta value for each possible pair-wise population comparison, one at a time against a panel of “parent” samples. Screened. Parental samples are samples from areas of the world that are relatively uniform, such as Niger or Congo for sub-Saharan Africans, Southern Mexico for Native Americans, China for East Asians, and Europe for Europeans.

  About half of the candidate AIMs proved not very useful because their actual delta values were not as high as expected from public database allele frequencies (some were not even SNPs, or Could not be assayed using the current platform). Sequences validated as true AIMs such as those illustrated herein are mixed mappings, inferring individual ancestry proportions, and inferring population group mixture proportions, plus their ancestry information It was useful to screen the genome to identify markers for alleles that correlate with specific human traits through donation. Even though each candidate AIM was initially selected from a public database based on rough population structure differences (i.e., continental populations), many of them separated human subgroups from larger groups throughout human evolution. Gave them a wealth of opportunities for genetic drift, founder effects, and natural selection that acted to either fix or eliminate their sequences, and to see that they possess information at a fine level of structure. It was issued.

  The sequences are shown in the sequence list from 5 ′ to 3 ′ (from left to right), and generally but not necessarily at nucleotide position 50 from the 5 ′ end for SEQ ID NOS: 1-331 Contains SNP (SEQ ID NO: 56, excluding position 26). Polymorphisms are indicated by the IUB symbol, S = C / G, Y = C / T, R = A / G, K = G / T, W = A / T, etc. As such, the disclosed sequences (SEQ ID NOs: 1-331) sample information about the target (ie, polymorphism) to be examined, plus the SNP (ie, measure the genotype of the sample) ) And amplification primer pairs, as well as information for preparing hybridization probes. Furthermore, the disclosed sequences can be used to scan public databases to identify additional upstream and downstream nucleotide sequences, if desired.

This panel of markers is extremely powerful for estimating the mixing ratio with accuracy in population samples (standard error is typically between 1% and 5%). In addition, AIM provided an affordable estimate of the ancestor of the individual, suggesting that equivalent accuracy could be obtained using more markers (confirmed in Example 2). Two independent methods of estimating an ancestor of an individual were used, maximum likelihood estimation (MLE) (Chakraborty et al., 1986) and Bayes' method using the program STRUCTURE (Pritchard et al., 2000); both The values obtained by the method were highly correlated when individual ancestry estimates were compared in terms of percent African genetic contribution in an African American sample from Washington, DC (R ² = 0.9836 ). These markers are excellent for measuring whether a sample from a mixed population has a population structure. As discussed below, the process of mixing is significant for the population and, as a result, a large number of false positive results (genetic structure, not by physical linkage of the marker to the disease-causing gene. This ability is important for mixed mapping applications because it can significantly increase the risk of misleading mapping results.

  This study confirms that AIM can be identified using the disclosed method, and that of 32 AIMs that can be applied towards the final goal of editing a panel of approximately 1,000 AIMs spanning the entire human genome. Provide a panel. Candidate AIMs were obtained by screening SNP allele frequency data generated by the SNP Consortium (TSC). Six sites, including Sanger Centre, Celera Genomics, Washington University, Orchid Biosciences, Motorola, and Whitehead Institute, as of 2003, each from three groups (African American, European American, and Asian American) Using a central collection of 42 individuals, we generated allele frequencies for 60,000 SNPs located throughout the genome. This database is freely available to researchers (see, for example, the URL “snp.cshl.org” using the Hypertext Transfer Protocol (“http”)), but used to provide the results. And accordingly demonstrated the usefulness of the source for editing the genome-wide panel of AIMs.

  This study focused on the accuracy of the SNP database and the number of candidate SNPs present. Regarding database accuracy, each group included in the SNP consortium took a different approach to creating data. As such, we addressed the initial concerns regarding how data can be combined. Because the genotyping approach was different for each group, it was necessary to address the problem of confirmation bias that could affect the data in a particular group differently. For example, the majority of groups generated their allelic frequencies after sequencing a subset of the TSC diversity panel, and then assigned these markers to the larger group of 42 individuals from three populations. Scored. The Washington University group took an approach where pooled sequencing across the region was performed and allelic frequencies were calculated for the variable positions found during this attempt. The Orchid group did not use sequencing, but instead started with a locus from the TSC SNP database known to be polymorphic. Given such differences, systematic characterization was made to the extent that different deflections, if any, may have influenced the results.

One approach to systematically characterize such potential bias was to compare allelic frequencies for loci genotyped by more than one group. As expected, frequency data obtained by different groups that varied around the 45 ° line, but showed a limited degree of allelic frequency bias introduced by different genotyping and confirmation strategies There was a general arrangement in (R ² = 0.8762). The next step in testing the accuracy of these data is to categorize the data by site and perform a pair-wise comparison, which is more specific for a particular site with an allele frequency estimate that deviates more when compared to other sites. Allows identification.

  With regard to the number of candidate AIMs, it was also important to determine how many of the 60,000 SNPs characterized by TSC are useful for mixed mapping. Since it may eventually be useful to compile a panel of approximately 1,000 markers that show large frequency differences between relevant population groups (Fst> 0.4), what percentage of the available markers is the desired feature It was important to evaluate whether Candidate AIMs were based on the recommendations of McKeigue et al. (Supra, 2000). Markers with information available to African, Asian and European populations, cumulative percentage of markers in each Fst category (0-1 at 0.05 interval) and total number of candidate AIMs for each possible comparison. The distribution of pairwise Fst from the TSC allele frequency project was as follows: Asian-European (556 candidates AIM / 25,110 total SNPs; mean Fst = 0.0720); Asian-African (1026 Candidate AIM / 25,578 total SNPs; mean Fst = 0.0886); and European-African (1306 candidate AIM / 30,103 total SNPs; mean Fst = 0.0861). As such, screening has shown that about 2-5% of the markers can be useful for mixed mapping.

  A mixed geographical pattern in a mixed population in the United States, particularly African Americans and Hispanics, was the subject of the first study. A mixed proportion of more than 18 African-American populations was characterized and a map was created showing estimates of the European genetic contribution to African-Americans from several different geographic regions in the United States . European mixes ranged from 3.5% in South Carolina galas to 22.5% in New Orleans (eg, 18.8% in Chicago; and 16.4% in Houston). Most of these estimates were obtained using the first panel of AIM that gave 10 pieces of information. The observed distribution was interpreted in terms of known historical and demographic events that played an important role in African American history (Parra et al., Supra, 1998; Parra et al., Supra, 2001. reference). These data allow the application of mixed mapping to identify genes involved in complex diseases. Mixed mapping is more appropriate in populations that show a high degree of mixing, therefore populations such as Gala (3.5%) and Jamaicans (6.6%), whose genetic contributions of Europeans were very limited It is expected that it may not be suitable for this type of analysis.

  In preliminary studies using mitochondrial DNA (mtDNA), African-Americans are low but detectable, consistent with self-reported native American ancestry often spoken by African-American individuals It was observed to have a genetic contribution of Native Americans. We identified 30 AIMs that gave information about African / Indigenous American contrasts and 19 AIMs for European / Indigenous American contrasts (see Table 1; see also Shriver et al., Supra, 2003), but 3 The presence of indigenous Americans in two African American populations was examined using nuclear DNA markers. Consistent with mtDNA estimates (only providing “maternal” contribution information), evidence of low indigenous American genetic contribution was detected in each of the African American samples (Washington, DC, 6%; from London) African Caribbean, 5%; and Bogarousa, Louisiana, 6%).

  For mixing in Hispanic, relevant European, Native American and African contributions were estimated in Spanish American samples from San Luis Valley CO. A 59% European mix, a 35% Native American mix, and a 6% African mix were observed in this sample and were in good agreement with the estimates previously described for the Mexican ancestry population (Chakraborty et al., Supra, 1986; Hanis et al., Supra, 1991; Tseng et al., Amer. J. Phys. Anthropol. 106: 361-371, 1998; Collins-Schram et al., Supra, 2002). Further characterization of the mixing in additional samples from Mexico and two samples from Puerto Rican ancestry Hispanics (New York and Puerto Rico) was performed as shown in Example 2.

  Samples of individuals of European ancestry (N = 199) currently living in State College PA were also analyzed. The genetic contribution in this sample was mostly of European origin (91%) and there was evidence of some African (3%) and indigenous American (6%) effects. These results are summarized in Figure 4 using a triangle plot, clearly showing the difference in average mixing levels among European Americans, Spanish Americans and African Americans. The triangle plot shown in Figure 4 shows the mean mixed estimate for a particular sample; the distribution underlying the ancestry of individuals is complex, and different individuals are widely distributed among African, European, and Native American ancestry Value (not shown). In African Americans, most individuals showed predominantly African genetic contributions, but some showed relatively high European contributions and also to a lesser extent indigenous American ancestry . European Americans were more closely packed near the pole in response to high European contributions, with few showing evidence of indigenous American and African ancestry. Spanish Americans showed the highest variability in individual ancestry, as expected when assuming the high mixing levels observed in this sample.

  In particular, individuals showed a full range of Europeans and Native American ancestry (from 100% Europeans to 100% Native Americans), and relatively lower African genetic contributions were also apparent in some individuals . Some of the variation observed in the ancestry of individuals was likely due to stochastic errors due to the limited number of markers used to infer ancestry. Thus, 20-32 markers used in the illustrated test detected the ancestry of the individual, but the standard error of estimation was quite high; increasing the number of AIMs It is expected to increase accuracy (see Example 2). Another component of variation in individual ancestry was due to true differences in ancestry between individuals. The striking correlation in individual ancestry values obtained by two completely independent methods, ML and STRUCTURE (discussed above), underlies that this panel of markers is unique to these populations It shows that an ancestor pattern of an individual can be captured. As disclosed below, controlling variation in an individual's ancestry allows avoidance of false positive results.

  The effects of mixing dynamics on population structure and linkage disequilibrium (LD) were investigated. The importance of the mixed model (hybrid sequestration model versus continuous gene flow model) with respect to the population structure and LD that occur during the mixing process has been described previously (Pfaff et al., Supra, 2001) and the level of population structure in the mixed population. Two methods to quantify are presented. Population structure is a key aspect of mixed mapping and any genetic association analysis in mixed populations. This issue was explored in African-Americans, Spanish-Americans and European-Americans using more informative markers than previously available.

  The existence of the structure was evaluated in two different ways. First, the observed number of significant associations was compared with the number expected at the 5% significance level among unlinked markers, and second, the average correlation of individual ancestry Estimated using two subsets. Consistent with previously reported data, the African-American population from Washington, DC is reflected by a very high number of significant associations between markers that are not linked more than expected by chance, Significant genetic structure was shown (10.5% vs. 5%, FIG. 5A). A very strong association was observed between markers as far as 24 Mb (AT3 to F13B, G = 15.21, p <0.0001), and clear evidence that these significant associations were caused by the mixing process Provided. The related alleles are always combinations of alleles that are frequent in the African population, the higher the difference in frequency, the more frequent association was observed between markers: The highest frequency difference between Africans and Europeans, but significantly associated with 9 unlinked markers. Thus, the mixing process in this African-American population, representing 17% European ancestry, resulted in a strong association between markers that showed a high frequency difference between African and European populations. These associations are significant both when the markers are linked and when they are not linked, but linked markers tend to show higher G values than unlinked markers, and true Linkage associations have been shown to be distinguishable from associations by genetic structure, as previously demonstrated (McKeigue et al., Supra, 2000). Interestingly, a significant association is found between European and African populations, not between markers that show high frequency differences between African and Native American populations, or between European and Native American populations. Observed between the indicated markers. This result was likely due to the low indigenous American ancestry (6%) observed in this sample and was insufficient to produce a detectable association by the mixing process in such a small sample.

Another line of evidence demonstrating the high level of genetic structure present in this African American sample was a significant correlation between independent estimates of individual ancestry using different subsets of genetic markers. The average correlation for 100 random selections of independent subsets of markers was r = 0.40, p <0.0001 (FIG. 5B). Genetics observed in Washington DC African American and other African American samples (Jackson MI and Lowland, South Carolina, Pfaff et al., Supra, 2001) analyzed with a more limited set of markers Structure and LD patterns indicated that the best model to explain the mixing process in these populations was a continuous gene flow model. Additional support for this model came from a strong correlation between D ₀ (the first expected association between markers caused by the mixing process) and D _t (the current association between markers). As shown using computer simulations (Pfaff et al., Supra, 2001), a positive correlation between D ₀ and D _t is expected in a population following a continuous gene flow model, and in fact this result is an African American population Observed in. In the population following the hybrid sequestration model, no significant correlation between D ₀ and D _t is expected.

Analyzed Spanish-American samples from San Luis Valley showed less genetic structure than any of the African-American population. The number of significant associations observed between unlinked markers was only slightly higher than expected at the 5% significance level (7.3% vs. 5%, Figure 5A). This result is interesting given that the Spanish-American population was much more mixed than African-Americans, and would be expected to show much more structure under the same model of mixed dynamics . The correlation of individual ancestry estimates based on independent markers was significant, but much lower than that observed in the African American population (r = 0.11, p <0.0001, FIG. 5B). Also, in the San Luis Valley sample, there was no correlation between D ₀ and D _t , in contrast to the results observed in African Americans. These results show that mixed dynamics (the process by which the population is formed and evolved) are different in African-American and San Luis Valley populations, the former being more similar to the continuous gene flow model than the latter. It has been demonstrated. Of course, other Hispanic populations may exhibit mixed kinetic patterns different from those observed in San Luis Valley.

  As expected from the lower mixing levels observed in European Americans, this sample from State College PA had no evidence of genetic structure due to mixing (FIGS. 5A and 5B). The number of significant associations between unlinked markers was similar to the value expected by chance, and there was no correlation between the ancestry estimates of individuals in an independent subset of markers (p = 0.149, NS).

  These results demonstrate that the use of selected genetic markers (AIM) allows analysis of the dynamics of the mixing process and the effects of this mixing process on LD patterns in a mixed population . Few positive results are expected in a mixed population with a mixed process similar to the hybrid sequestration model (after initial mixing, followed by independent evolution of the mixed population without further genetic contribution of the parent population) ( (Recall that false is related to “gene hunters” that search for genes that cause traits through LD or linkage, not those who strive to develop classification tools). In a mixed population that more closely resembles a continuous gene flow model (generational continuous genetic contribution from one of the parent populations to its mixed population), LD can spread over much longer distances. Problems with expected and false positive results will occur. Fortunately for gene hunters, the information conveyed by AIM can control the genetic structure and minimize false positives. Examples are provided below that demonstrate how such control can be achieved using skin pigmentation as an appropriate statistical method and model phenotype.

  Skin pigmentation and individual ancestry were examined in African American and Spanish American samples. As previously demonstrated, the genetic structure resulting from mixing can be effectively controlled, and linkage linkages can be distinguished from false linkages due to genetic structure using appropriate statistical tests ( McKeigue et al., Supra, 2000). In this study, the same method was applied in the study of skin pigmentation in two mixed samples (African American from Washington DC and Spanish American from San Luis Valley). Information on skin pigmentation was collected for each individual in both studies, subjects were genotyped for a panel of AIMs, and the ancestry percentage of individuals was calculated using the maximum likelihood method (Chakraborty et al., Supra, 1986). . Individual ancestry (% African or% Native American) was plotted against melanin index (African) or skin reflectance (Indigenous American) for each individual. Some of the AIMs that showed high differences in frequency between parental populations were also candidate genes for pigmentation.

In African-American samples, a strong and highly significant correlation (R ² = 0.1879, p <0.0001) was observed between the ancestry of the individual and the melanin index, which measures skin melanin content. Individuals with darker skin on average had higher levels of African ancestry. An individual's ancestry estimate is based on 21 markers and is therefore subject to a relatively high variance, thus accounting for at least some of the variations observed in the graph. An interesting feature of these results was the apparent decrease in variance observed in moving from right (more African ancestry) to left (more European ancestry). This result is consistent with the high level of variability in skin color found in African populations compared to Europeans. The high correlation observed between individual ancestry and skin pigmentation is due to the population structure typical of African-American populations, and the genes used to measure parental population differences contained within this relationship Can be related to a limited number of

A similar plot was prepared for a sample of San Luis Valley. Individual ancestry estimates using 15 Native American / European AIMs were plotted against pigmentation levels measured by the percent of light reflected through the PHOTOVOLT 670 green filter. Since skin pigmentation was measured in different ways (absorbance vs. reflectance) in these two studies, the trend observed when shown graphically is reversed. In the Spanish-American sample, the correlation between individual ancestry and skin color was also significant (R ² = 0.0481, p <0.001), but lower in the African-American sample and probably present in this sample This is probably due to the reduced genetic structure.

  Tests were performed for differences in average pigmentation levels by genotype for AIM typed in the African American population sample discussed above. The AIM panel included three candidate genetic markers, OCA2, TYR and MC1R. The analysis was performed in three alternative methods: the first method (ANOVA) that does not take into account the individual's ancestry estimation; the second after conditioning to control the influence of the individual's ancestry that excludes the locus under consideration Method (ANCOVA / IAE minus marker); and a third method (ANCOVA / IAE) using complete individual ancestry estimation for conditioning. As shown in Table 2, 8 out of 21 markers (38%) showed a significant difference (p <0.05) among the 3 genotypes, and 2 of the 4 candidate gene markers (OCA2 and TYR). When using an alpha level of 0.05, only 5% of the tested markers were expected to yield significant results. As such, the discovery of a significant difference of 38% indicates that the population structure is associated with both ancestry and pigmentation (Pfaff et al., Supra, 2001; Parra et al., Supra, 2001).

  One way to remove the effects of population structure is to test for differences that are a condition of individual ancestry estimation (IAE). When complete IAE was used to establish the condition (ANCOVA / IAE), only one locus, OCA2, the human P gene showed a significant average difference between genes. When a less conservative conditioning approach was taken (ANCOVA / IAE minus marker), where the locus under consideration was excluded from the individual's ancestry estimates, there were four significant results: OCA2, TYR, FY, and SGC30055 It was.

  Bayesian complete probability models for mixed and marker genotypes have also been established (McKeigue et al., Supra, 2000). Scoring tests for linkage test for independent association between the number of European ancestry alleles at each locus and pigmentation, one at a time in a regression model that includes individual ancestry (estimated from marker data) Based on what to do. The one-sided probabilities for scoring tests are shown in Table 2, with three loci showing evidence of linkage to skin pigmentation at an alpha level of 0.05 (OCA2 (p = 0.005), AT3 (p = 0.027) , TYR (p = 0.033)}. To confirm these results, other markers that give information about ancestry in OCA2 are to be identified and analyzed by scoring methods. The agreement between these ANOVA results and Bayesian mixed mappings is encouraging, and both methods will benefit from the addition of new unchained AIMs and increase the accuracy of individual ancestry estimations.

  Spanish-American samples from San Luis Valley CO were also analyzed for linkage and association using Bayes and ANOVA methods (Table 3). This analysis included 442 individuals typed for 15 marker loci that give information about ancestry (two SNPs in the DRD2 gene are treated as one locus). The CYP19E2 marker (MYO5A, located near the pigmentation candidate gene) showed strong evidence for linkage to ethnic differences in skin pigmentation. However, this result should be interpreted with caution because, as long as some closely linked markers that give information about ancestry are not used, the test for linkage is an ancestor-specific allele. It is because it is not strong against the wrong specification of frequency. SNPs around MYO5A can be analyzed to confirm these preliminary results.

¹ マーカーは、検定に用いられる祖先情報提供マーカーを示す。太字のイタリック体で示されたマーカーは色素形成についての候補遺伝子である(すなわち、OCA2、MC1R、TYR)。
² 「デルタ」(δ)は、アフリカ人とヨーロッパ人集団の間の対立遺伝子頻度差である。
³ 性別が唯一の共変動である、分散分析有意水準
⁴ 検定される遺伝子座が共変動として除外された個体の祖先推定(M)を用いる一元ANCOVA分析についての有意水準。
⁵ Mがすべての21個のマーカーに基づくことを除いて、3と同じ。
⁶ ベイズの混合マッピング片側確率。 Table 2. Tests for the effect of single locus genotypes on pigmentation in African American samples

¹ marker indicates an ancestral information providing marker used for the assay. Markers shown in bold italics are candidate genes for pigmentation (ie, OCA2, MC1R, TYR).
² “Delta” (δ) is the allelic frequency difference between African and European populations.
³ gender is the only co-variation, analysis of variance level of significance
Significance level for one yuan ANCOVA ^{analysis 4-validated} the loci used ancestry estimation (M) of individuals were excluded as covariate.
Same as 3, except ⁵ M is based on all 21 markers.
⁶ Bayesian mixed mapping one-sided probability.

¹ マーカーは、検定に用いられる祖先情報提供マーカーを示す。太字のイタリック体で示されたマーカーは、色素形成についての候補遺伝子の中に、または近くにある(すなわち、TYR-192、およびMYO5A近くのCYP19E2)。
² 「デルタ」(δ)は、先住アメリカ人とヨーロッパ人の親集団の間の対立遺伝子頻度差である。
³ 性別が唯一の共変動である、分散分析有意水準
⁴ 検定される遺伝子座が共変動として除外された個体の祖先推定(M)を用いる一元ANCOVA分析についての有意水準。
⁵ Mがすべての15個のマーカーに基づくことを除いて、3と同じ。
⁶ ベイズの混合マッピング片側確率。 Table 3. Tests for the effect of single locus genotypes on pigmentation in Spanish-American samples

¹ marker indicates an ancestral information providing marker used for the assay. Markers shown in bold italics are in or near candidate genes for pigmentation (ie, TYR-192, and CYP19E2 near MYO5A).
² “Delta” (δ) is the allelic frequency difference between native American and European parental populations.
³ gender is the only co-variation, analysis of variance level of significance
Significance level for one yuan ANCOVA ^{analysis 4-validated} the loci used ancestry estimation (M) of individuals were excluded as covariate.
Same as 3, except ⁵ M is based on all 15 markers.
⁶ Bayesian mixed mapping one-sided probability.

Pairwise population comparison of SNP discrimination Table 4 shows the results of genotyping and statistical analysis demonstrating several different but important points (the sequence for each AIM in Table 4 uses the marker number, Can be found by reference to Table 6).

主張されたリストからの精選されたAIMについての「デルタ」(δ)値が示されている。
AIM固有識別名は最後の列に示されている。
太字(影がついていない)の番号をもつセルは良いδ値を示す；太字の番号をもち影がついているセルは極めて高いδ値を示す。
AF-アフリカ人、CT-ヨーロッパ人、EA-東アジア人、SA-南アジア人、ME-中東人、PI-太平洋諸島系、AI-先住アメリカ人。 (Table 4) Pairwise group comparison of SNP distinction

“Delta” (δ) values for selected AIMs from the claimed list are shown.
The AIM unique identifier is shown in the last column.
Cells with bold (no shadow) numbers show good δ values; cells with bold numbers and shades show very high δ values.
AF-African, CT-European, EA-East Asian, SA-South Asian, ME-Middle Eastern, PI-Pacific Islander, AI-Indigenous American.

  First, Table 4 is derived from screening hundreds of candidate AIMs electronically selected from public databases, and thus only a small number of candidate AIMs from public databases are true AIMs. Proved that. As discussed above, the public SNP database was electronically screened to find good candidate AIMs (frequency data are provided for three “race” groups, so The level of mixing is unknown)). Second, 384 individuals of various continents and BGA origin are genotyped at each of these sites: 70 African samples collected from Nigeria and the Congo, 65 collected from Northern Europe European samples, 70 East Asian samples collected from recent immigrants to San Francisco, CA; 35 Middle Eastern samples collected from Turkey, 35 South Asian samples collected from India, and 25 Pacific Islander samples collected from the Philippines and US Samoa.

  Sampling of data for approximately 70 AIMs that passed the screening process from 175 candidate AIMs screened is shown in Table 4. The delta (δ) value is enough to allow the sequence for the polymorphism to predict membership in one or the other group; ie how different the two populations are with respect to the sequence in this polymorphism This is the scale. δ values are shown for 69 of 175 AIMs; the remaining 105 had a δ value of 0 for each of the pair-wise population comparisons and were therefore not true AIMs. AIM1068 in Table 4 is representative of the type of failure (There are several AIMs in Table 4 that have zeros across all pairwise comparisons-zero across population pairs, but these are Because it gives information about groups not shown in the table). This result confirms that most of the candidate AIMs selected from the public database are not true AIMs, and highlights the value of the present invention.

  A relatively large investment in genotyping and analysis required identifying which candidate AIMs were true AIMs. This process can be bypassed, for example, simply genotyping a sample in 100 candidate AIMs and extracting data from what proves to be truly AIM, but genotyping is an expensive method As such, the wasteful burden of making it makes the inspection economically impractical. In order to develop an economical and practical test for the proportion of ancestors, the test must ask a number of true AIMs. Many publicly available candidate AIMs provide some information, but allele frequencies in public databases are not very reliable, for example low samples Not true AIM because of size. The frequency of true (i.e. validated), correctly characterized (i.e. population-specific frequency is known with certainty) in random collections of SNPs is expected to be about 5% However, the true AIM frequency in the selected set of candidate AIMs is about 50%, and after proceeding as disclosed herein, the true AIM frequency in the collection of SNPs is 100% It is.

  Second, the results in Table 4 demonstrate that some AIMs are suitable for African-European decisions, and other AIMs are suitable for indigenous Americans-African decisions. Although selected based on European / African / Asian allele frequency differences, some AIMs provide good discrimination of other groups such as Pacific Islanders, South Asians and Middle Easterns. This type of information can only be known by genotyping in a larger sample, and testing for ancestral proportions must go through this stage to be accurate (e.g., testing If you only worked in the third dimension given by publicly available data-Europeans, Africans and Asians-for example, the results obtained for Hispanic would have been ambiguous. The panel of SNPs in Table 4 provides a well-balanced mix of AIM with decisive power for each of the possible pair-wise comparisons for the seven population groups, and this panel constitutes a good test for ancestry proportions It seems to do. Data for South Asians, Middle Easterns, and Pacific Islanders do not exist in public databases and were therefore created for these studies. In comparison, one attempt to develop a test for the proportion of ancestors in 7 dimensions by simply sequencing on candidate AIMs selected randomly from a public SNP database (ie without selection through data generation) Although certain pairs of populations are difficult to determine (e.g., South Asians and Europeans that make up a larger Indo-European group combined by a common language parent), those in Table 4 It is necessary to edit a group of thousands of SNPs to obtain such a panel.

  The results obtained using the AIM panel shown in Table 4 are shown in Table 5. The currently disclosed algorithm (Example 6, see Table 12) is used to calculate the proportion for a group of 96 individuals who reside in the southeastern United States and describe themselves as Caucasian. Used.

祖先のパーセンテージは各群に関して与えられている：EUR - インドヨーロッパ人、NAM - 先住アメリカ人、EAS - 東アジア人/太平洋諸島系、AFR - アフリカ人。各個体についての自己申告された人種(SELF)、彼らの母親(M)、父親(F)、母方の祖母(MGM)、母方の祖父(MGF)、父方の祖母(PGM)、父方の祖父(PGF)および彼らの出生国が示されている。 (Table 5) Proportion of ancestors for a number of self-reported Caucasians

The percentage of ancestry is given for each group: EUR-Indian European, NAM-Native American, EAS-East Asian / Pacific Islander, AFR-African. Self-declared race (SELF) for each individual, their mother (M), father (F), maternal grandmother (MGM), maternal grandfather (MGF), paternal grandmother (PGM), paternal grandfather (PGF) and their country of birth are shown.

  The results in Table 5 show that most people who describe themselves as Caucasians really use BGA testing, i.e. using the marker panel and algorithm in Table 4 (Example 6, Table 12). Demonstrates having most Indo-European ancestry, as measured. Approximately 40% of these Caucasians were measured as 100% European ancestry without mixing, while 60% showed detectable mixing. As a critical assessment, the percentage obtained can be compared to a self-reported blend. For individuals in Table 5 claiming that all of their parents and grandparents are unmixed Caucasians ("ca" across all columns), 90% or more European ancestry is found The rate is higher for those who report no mixing to their pedigree (55%) than those who report mixing to their pedigree (35%). The fact that half of the people reported no mixing in their pedigree is how most people don't know about their BGA, at least in an anthropological name compared to geopolitics. Show.

  The public database used a small number of samples for each of the three groups (African American, European and Asian). Therefore, the actual allele frequencies for the claimed SNPs from public databases were uncertain and were measured with accuracy only from this study. Furthermore, the use of African Americans as parent groups is prone to error because they are mixed populations (between Africans and Europeans), as disclosed herein. The best way to find SNP markers that are useful in this method is to genotype a large number of samples from the world's major BGA groups for all of the minor allele frequencies that are clearly different between at least two of the groups And calculate the δ values and rank them.

Family Tree The BGA method of using SNP (AIM) to measure the proportion of ancestry was applied to the examination of the family tree of the father, mother and their three children (see Figures 6 and 7). The men are mostly European and his wife is Mexican, so if the test is accurate, their three children plot somewhere between that man and his wife This is exactly what has been observed. Three of the father's grandparents on the paternal side of the family were relatively pure Greek / European, but one was almost pure Cherokee. All four of his grandparents on his mother's side were a mixed European. Plots for the mother (Figure 6B) and father (Figure 6A) are shown. By drawing a line from the vertex that bisects the side of the triangle opposite the vertex (the vertex represents 100%, the triangle side represents 0%), the father was about 85% European, 11% Native American, 4 It can be seen as an African (he had no detectable East Asian, South Asian or Pacific Islander ancestry). 11% Native Americans appear to have come from his father's grandparents. If you know that 7/8 of his great-grandparents were European mixed and 1/8 were indigenous Americans, the expected percentage was 12% (1/8), resulting from the test The data match well. The mother is from Mexico and is Hispanic. As previously discussed, Hispanic is a mixed population with European and Native American contributions. Mothers can be seen here with 11% European, 76% Native American and 13% African pedigree (no detectable East Asian, South Asian or Pacific Islander ancestry).

  Each of the three children received a chromosome from each of their mother and father, so they should each plot somewhere between him and his wife. Using that method, the children are plotting as expected (Figure 7). As expected, it is clear from these results that the child's point estimate is between that of the parents. From these results, child No. 1 (Fig. 7A; 80% European, 18% Native American, 2% African) has a proportion of ancestry more similar to father than mother, and child No. 2 (Fig. 7B; 61 % European, 31% Native American, 8% African) and Child 3 (Figure 7C; 54% European, 37% Native American, 9% African) are about halfway between the two parents Has the percentage of the ancestor of a point.

  Each child receives one chromosome from the mother and one from the father, although the mother has mostly indigenous American pedigree chromosomes, some still have European and African types, Although fathers have mostly European-type chromosomes, children are different because some are still indigenous-Americans. When a child is conceived, he or she receives a chromosome from each parent, but which of the two a child receives from a mother is random (ie, an “independent combination”). Since some of the mother's chromosome pairs have members of pairs that include the European type, some children receive the “European type chromosome” and others do not. From this study, it is clear that child 1 (Figure 7A) received more of the mother's European chromosome than the other two children. Thus, each child is a 50/50 mixture of mother and father chromosomes, but the proportion of their ancestors is their unique and random function of their parents.

Example 2
Ancestor's four-way mixed estimation This example demonstrates that a four-way mixed BGA test provides the same results obtained using a three-way BGA test.

  As indicated above, biogeographic ancestry (BGA) is the genetic component of race. Because socio-cultural and geopolitical metrics for measuring humanity are human-made, non-natural, constructs, their use in genetic research makes it difficult to control the genetic structure of a population And may obscure an important correlation between BGA and human biology. This example provides methods and compositions for accurately measuring genetic structure within an individual. The human genome has drawn interest in candidate ancestral informative markers, which have been validated in an ultra-high throughput genotyping platform and used to establish parental allele frequencies. Using 71 of the most informative AIMs that cover most of the chromosomes (Table 6), the human population is divided into four main continental population groups (sub-Saharan Africa, East Asian, Indo-European and indigenous) The MLE method was used to determine individual BGA mixing ratios and their associated confidence intervals. As disclosed herein, self-reported population affiliations were almost completely correlated with the majority of BGA population affiliations measured for one sample of 2,024 international samples. BGA mixing results were surprisingly frequent and were generally consistent with anthropological and geopolitical history when observed. The created mixing ratio is traced in the family tree in a manner consistent with the law of independent combination, and by simulation, markers related to determining group affiliation are independent within the scope of our algorithm. It was revealed that it worked. The test was surprisingly robust because a large number of high delta markers were used; a reasonable level of simulated allele frequency error that could be caused by biased parental sampling was measured in the measured BGA percentage There was no significant effect. These results demonstrate that BGA mixing can be reliably measured from DNA samples.

Sample collection:
Parental Samples—To establish parental allele frequencies, 100 relatively uniform offspring of four roughly defined human population groups were genotyped. These four groups are compared to continental regions in which the population is sub-Saharan Africa (sub-Saharan Africa), Europe and Middle East (Indo-European), North / South America (Indigenous Americans), and East Asia (East Asians) Corresponds to the unification of simplified human genealogy that goes back to the point where it was isolated. In essence, extrema of existing populations are taken by physical characteristics and known human migration patterns, and the anthropology has been simplified by assuming that because the rest of the population is due to these extremes This is because it reveals physical features along the defined continuum, and all of human society was born out of intrusion and mixing between these four main continental groups.

  Collection efforts focused on individuals living in each area and having a relatively uniform affiliation for each group's offspring by a self-described “race”; each subject was associated with each group's offspring Reported a uniform physical affiliation for the group. There is no de facto ultimate referee of BGA affiliation or race, but collecting without considering ethnicity is the frequency for a given group when systematic mixing is a function of ethnicity. Systematic errors can be introduced into the estimation. Where possible, attempts were made to collect from the widest possible range of ethnic groups within each parent group, and the mix within each parent sample was expected to be balanced. It would be better to collect from individuals of known uniform affiliation, but if the sampling was not biased with respect to the mix or other population structure, the existing mix in the sample that is practical to collect would be Rather than introducing systematic deflections, they tend to reduce the power of inspection. The presence of Hardy-Weinberg equilibrium for each marker within each BGA group was relied upon as an indication that a reasonably good sample was obtained. Sub-Saharan African samples were collected in Nigeria and Congo, Africa; European samples were collected from various locals in the United States; East Asian samples were collected from Japan and China; and Native America Human samples were collected from “Nativos” in remote areas of southern Mexico. All samples were collected under IRB guidelines for genetic studies of human population variability.

  Experimental Sample-After reading and signing an approved IRB consent form, the subject completed a biographical questionnaire and provided either buccal application or 4 ml of blood. In the questionnaire, subjects will identify themselves, their mother, father and maternal and paternal grandparents as “African”, “American Indian”, “Asian”, “Caucasian”, “Hispanic” or “Other”. It was described including the option of reporting "I don't know" about each family member as belonging to the group. Digital photographs were taken of some of the subjects; explicit permission was obtained from those subjects to whom the photos were provided. DNA was extracted from circulating lymphocytes or buccal applications using a commercial kit (Qiagen) and a primer extension protocol using a 25 K SNP stream ultra high throughput (UHT) genotyping system was used ( Orchid Biosciences).

Biogeographical ancestry (BGA) estimation software program uses the algorithm of Hanis et al. (Supra, 1986) to determine the maximum likelihood estimate (MLE) of an individual's BGA mixture using a multidentate AIM genotype. Based on (Example 6; see also Table 12). The delta (δ) value is a representation of the marker ancestry information provision (Dean et al., Supra, 1994). For biallelic markers, the frequency difference (δ) is equal to p _x -p _y (which is equal to q _y -q _x ), where p _x and p _y are the frequencies of one allele in populations X and Y , Q _x and q _y are the other frequencies. To test the deviation from independence in allelic status within and between loci, an MLD exact test was used (Zaykin et al., Genetica 96: 169-178, 1995).

  The collection of 71 AIMs used in this example is cumulative within each of the six possible pairs of four-dimensional (sub-Saharan African, Native American, Indo-European, and East Asian) problems. It was chosen to maximize the δ value and minimize the difference in cumulative δ values between each of the pairs. The algorithm reverses the population-specific allele frequency (mainly for computational convenience and 4D) to obtain a likelihood estimate of proportional affiliation corresponding to the multilocus genotype using 3 groups at a time. Because mixing is likely to be relatively rare). For example, the likelihood of 100% Indian Europeans, 0% Native Americans, 0% East Asians is calculated, then 99% Indian Europeans, 1% Native Americans, 0% East Asians are calculated next And so on until all possible Indo-European, Native American and East Asian percentages are considered, and then the process is followed by all possible Indo-European, Native American and African percentages , And repeated for all possible Native American, African and East Asian proportions. The likelihood of the maximum value is selected as the maximum likelihood estimate (MLE).

  When plotting a single MLE on a triangle plot, the space where the likelihood is within 2x, 5x and 10x the MLE is bounded (if multiple MLEs are shown in a single triangle plot, These intervals are not plotted). For calculating MLE using all four BGA groups together, the procedure was performed in the same precision manner; instead of three possible three-way BGA combinations, only one four-way BGA combination is possible is there. All MLEs described in this example were calculated using a ternary calculation scheme. Alternative versions of this type of examination are possible, for example, using different AIMs and different parent groups corresponding to different coalescings of the anthropology, with respect to different anthropological time scales than those given here It is expected to provide meaningful results.

  When the database was screened, the SNP Consortium (TSC) contributed data on about 27,000 SNPs that are available in three populations (African American, European American and East Asian) did. This database was screened for candidate AIMs, ie SNPs with δ> 0.40 between any two of the four continental population groups (see Example 1; see also Shriver et al., Supra, 1997). Sub-Saharan African (AA), Indo-European (IE), East Asian (EA) and Native American (NA) parent samples were screened for each of the 200 candidate AIMs with the largest δ values, Of these, 71 were confirmed as true AIMs (ie, true SNPs) with a minor allele frequency greater than 1% and δ> 0.40 for at least one of the pair of groups. 71 AIMs are shown as SEQ ID NOs: 1-71; the top 100 candidate AIMs for the group pairs were as follows: EA vs. AA (SEQ ID NOs: 7, 21, 23, 27 45, 54, 59, 63 and 72-152); EA versus IE (SEQ ID NOs: 3, 8, 9, 11, 12, 33, 40, 59, 63 and 153-139); and IE versus AA (sequence Numbers: 1, 8, 11, 21, 24, 40, 172 and 240-331). That some AIMs identified by one pair-wise comparison could also be AIMs for a second pair-wise comparison (eg, SEQ ID NO: 59 was identified as a comparison of EA vs. AA and EA vs. IU) It should be noted that such an AIM is an exception. Furthermore, the majority of 71 AIMs are not in the list of top 100 candidate AIMs shown for any pair (but were in the top 200 candidate AIMs); They were not used because they did not fully genotype because of the SNP type of amplification parameters used for the illustrated platform, or for other reasons as disclosed herein .

AF-サハラ以南アフリカ人、CT-インドヨーロッパ人、EA-東アジア人、AI-先住アメリカ人。AIM固有識別名は、SNP配列のNCBI:dbSNPデータベースへの提出において著者らにより与えられたGenBankアクセッション番号と同様に示されている(AIM)。各ペアワイズ比較についてδ>0.40をもつAIMの数は、リストの一番下に示されている。 (Table 6) Pairwise δ values for 71 AIMs used for BGA inspection

AF-Sub-Saharan Africa, CT-Indo-European, EA-East Asian, AI-Indigenous American. The AIM unique identifier is shown as well as the GenBank accession number given by the authors in the submission of the SNP sequence to the NCBI: dbSNP database (AIM). The number of AIMs with δ> 0.40 for each pairwise comparison is shown at the bottom of the list.

  The 71 AIMs used in the illustrated panel spread across 21 of the 23 autosomal chromosomes (Figure 8), with the average chromosome containing 3 AIMs (see Table 6). ). Each has alleles in the Hardy-Weinberg equilibrium and is in linkage disequilibrium with each other, both in all four BGA groups considered together and within each BGA group. It was not issued. The software program used individual genotypes for these AIMs with a maximum likelihood algorithm (see Example 6, Table 12; see also Example 1). The use of 71 markers in this algorithm provides another example of “BGA inspection”.

  The BGA test was used to calculate the BGA mix ratio for the parent Native American, African and Indo-European samples used to construct the test. After calculating the mixing ratio for each sample, they were plotted in a triangular plot to allow the relative ratio of the ternary mixing to be represented in two dimensions. Because these are the same sample including the parent group and from which the population allele frequency was derived, they are expected to exhibit a relatively uniform BGA (i.e., low mixing), indeed, sub-Saharan Africa Human, Native American and European parent samples all showed a relatively uniform BGA (ie, they were plotted towards the appropriate vertices of the BGA triangle).

  The BGA test was then used to measure the BGA rate for 1,186 individuals of self-reported race (43 African Americans, 1,120 Caucasians and 23 Hispanics). 306 (26%) of the individuals showed uniform BGA (100% for any one group). 101 of 1,186 (8.5%) included> 5% BGA affiliation for 3 groups, and for these individuals, software modifications to perform 4-group calculations may be more appropriate, And the majority of samples showed that they had the characteristics of binary mixing. A significantly larger European mix was identified in African Americans compared to that of Nigerians (visualized as a variance of points away from the sub-Saharan African peak). In contrast, the Indo-European sample plotted as neatly at the top of the Indo-European population as the parent sample had, but a low level of Native American or East Asian mixing was not uncommon − Approximately 2/3 of the subjects are detectably included, but are generally such a low level of mixing. Hispanic subjects plotted with equal distribution along the Native American / Indo-European axis and knowledge that Hispanic was born about 500 years ago from a mix of colonial Europeans and resident Native Americans Matched.

  To determine if and to what extent the majority of the BGA results obtained were self-declared, for 2,048 individuals of the race with a self-reported BGA mix Most of the BGAs calculated and measured from the test were compared (in the sense of calculation) blinded to the self-declared majority of each individual. A very strong agreement was observed between most BGA groups measured by examination and most self-reported races (Table 7). Using the test, 1252/1252 self-described European Americans (US-born Caucasians) showed mostly Indo-European BGA. Of the 201 self-listed African-Americans, 191 show mostly sub-Saharan BGA, and the remaining 11 belong mostly to sub-Saharan BGA and mostly Indo-Europeans. As shown. Hispanics showed a roughly equal distribution in BGA between Indo-Europeans and Native Americans, consistent with the results observed in the triangle plot and the anthropological history of this group.

^＊ 小部分割合は先住アメリカ人であった。
^＊＊ 第二型として有意な先住アメリカ人
^＊＊＊ 小部分割合は東アジア人であった。 (Table 7) Comparison of most biogeographic ancestry with self-reported race

^{* A} small percentage was Native Americans.
^** Native Americans that are significant as the second type
^{*** The} small fraction was East Asian.

  Even if an unexpected result was obtained, such as the discovery that one individual from Southern Mexico was mostly European, a small part rather than no expected affiliation The affiliations were more or less consistent, and most of the affiliations (Indo-Europeans) seemed to be in light of the history of the region (which was colonized by the Spanish hundreds of years ago). In this particular case, the proportion of Native American ancestry was only slightly less than 50%. Only one terrible error was observed, for example, self-reported European Americans were mostly classified as East Asians. Most of the 11 self-described African Americans in the Indo-European BGA would seem to be a terrible mistake, but the results were equally consistent-each of these African-American samples , Showing roughly similar Indo-European / African proportions, suggesting mixing, and consistent with the historical tapestry of the area where the samples were taken, rather than inspection errors.

  The San Diego Police Department Crime Lab (SDPD) and the University of Central Florida National Center for Forensic Science (SDPD) to conduct a true blind test (as opposed to a test that is blind in the sense of calculation) The National Center for Forensic Science at the University of Central Florida (UCF) each deposited 10 buccal applications of numerically encoded identity for BGA testing. A BGA test was performed and results were returned to SDPD and UCF, each independently evaluating their results to reveal the self-reported population affiliation of the samples. Most percentages measured from the BGA mix test were consistent with self-reported group affiliations (see Table 8). Some of the samples were from individuals belonging to groups that were logically considered mixed--for example, Filipinos (SDPD2, SDPD3, Table 8), African Americans or Caribbeans (SDPD5, SDPD6 UCF7, UCF8, Table 8), Mexican American (Hispanic, SDPD8, SDPD10, Table 8) and Puerto Rican (UCF6, Table 8); significant mixing was detected for each of these samples. Still further, the type of mixture detected was in line with the anthropological history of the group to which it belonged. For example, sub-Saharan Africans, indigenous Americans and Indo-European pedigrees live in Puerto Rico, while East Asians are relatively rare, and test results for tested Puerto Rican individuals are: Showed Indo-Europeans and Native Americans, but did not show a mix of East Asians.

  In addition to determining the mixed MLE, the software program surveys the probability space, and within the range, the likelihood of proportional affiliation is two, five and ten times more likely than MLE to be the correct answer, Designed to define that space low (confidence contours, plotted on a triangle plot as a ring centered on MLE). One method to test the accuracy of the maximum likelihood algorithm for determining MLE and confidence contours is to analyze by replacing some of the AIM markers with genotypes for each with a “failure” reading When removed from, it was to observe whether and how much these values would change. For example, for a given sample genotype, if all of the genotypes for high δ markers for African / East Asian discrimination were replaced with “failure” or “no data”, then the correct test is It would be expected to show a distorted confidence contour only in this dimension of the triangle plot. Thus, the BGA test is mostly used to plot one sample of East Asian BGA with its associated confidence interval (Figure 9A), and then informative about Indigenous American vs. East Asian BGA All 24 of the markers in the test with δ were removed, and MLE and confidence estimates were recalculated with the remaining AIM (FIG. 9B). In the missing AIM recalculations, the trust circle is distorted dramatically from East Asians to the indigenous American BGA apex (Figure 9B), and as expected, good for East Asian / Indigenous American discrimination. This shows that a deficiency of AIM with a δ value resulted in a less reliable estimate along the East Asian / Indigenous American axis. Maybe the sample was mostly classified as East Asian, and the AIM on identifying East Asian / Indo-European and East Asian / African affiliations was left undisturbed, so the MLE shift itself Most of the uncertainty along the East Asian / Indigenous American axis was evident in the contour shift. Similar experiments with other samples and AIM produced similar results.

(Table 8) Blindfolded challenges in BGA testing by the San Diego Police Department and the University of Central Florida Forensic Science Center

  Five samples were genotyped and analyzed in separate cases to determine the reproducibility and consistency of BGA mix fraction measurements. One sample was selected from the majority of self-reported affiliations to European Americans, African Americans, Hispanics, and Asians, and a fifth sample was selected from the parent Native Americans. Except for the failed locus, the genotype at each marker in each individual is 100% consistent between runs, indicating that AIM is genotyping reliably. A 1-3% variation in BGA mixing ratio was observed throughout the run; simulation studies showed that the variation was attributable to these genotyping failures. This result indicates that the failed locus did not significantly interfere with the reproducibility of the BGA mixed measurements for the individuals, mostly for either BGA or mixed levels. From these simulations, if the sample is a minimum binomial mixture, the BGA test will tolerate approximately 10 locus failures and, for samples without mixing, a larger number of failed loci can be tolerated. Was measured (ie, the change in mixing percentage was less than 5%).

  Given the laws of family inheritance, the proportions for three generations were calculated from several pedigrees to determine if the BGA mixture proportion measured in the test was perceptible. A typical result is shown in FIG. 10, depicting the ratio obtained for a confirmed paternal (with STR test) family with a substantial European / Indigenous American mix. The first generation individuals were self-reported European Americans and were measured to contain a significant indigenous American mix on a BGA test, which was communicated to their sons and daughters in different proportions, independent combinations There was no contradiction with the law. The son's spouse was a Mexican-born Hispanic, measured to be 26% European / 74% Native American mixed. Each of their children roughly contained an intermediate level of indigenous American and Indo-European mixing between their parents and again was consistent with the law of independent combination. One of the sons was classified as having a small percentage of East Asian ancestry, but its level (4%) was within the established high confidence limit (about 3%) as discussed above. It was close. Other pedigrees tested (n = 8) showed similarly consistent results, indicating that BGA test results were perceptible within the pedigree for most BGA and mixed levels.

  Because the BGA test relies on parental allele frequency, the extent to which the mix ratio created by the test was influenced by the parental allele sampling bias was examined. A pair of groups was selected and AIM was associated with determining affiliation between these two groups (Table 6-selected the one with the highest δ value for this pair of groups) Allele frequencies were adjusted for each of these AIMs in one of the groups so that the δ value for AIM was reduced by 20% (for these two groups). In short, the power of the test was systematically reduced by 20% for the determination of affiliation between specific pairs of groups; this reduced test is called the ancestor 2.1EA / EU BGA test, EA / EU refers to the decline in discrimination between East Asian (EA) and Indo-European (EU) groups. Thirty-one samples were randomly selected and the genotype was run against the ancestor 2.1EA / EU BGA test in exactly the same manner as for the prototype (ANCESTRYbyDNATM 2.0), and the result was ANCESTRYbyDNATM Compared to that obtained in the 2.0 test.

  If the results from the ANCESTRYbyDNATM 2.0 test can be expected to be on the order of a few percent of the parent's allele frequency caused by sampling bias, but are highly sensitive to ancestry The 20% change introduced for 2.1EA / EU AIM should result in a mixing ratio that is substantially different from that of ANCESTRYbyDNA ™ 2.0. European / East Asian because the number of Indian European / East Asian mix observed was significantly greater than other types of mixes such as African / Asian mix or Native American / African mix Pairs were selected for the first study-ancestor 2.1EA / EU-and the BGA group pair with the lowest AIM count and cumulative δ value was Native American / East Asian, and the δ value for this pair was Changed in second test-ancestor 2.1NA / EA. The average change observed in the mixing ratio between ANCESTRYbyDNA ™ 2.0 and ancestor 2.1EA / EU was 1.4% (standard deviation 2.44%). For the Native American / East Asian pair, the mean change between ANCESTRYbyDNA ™ 2.0 and ancestor 2.1NA / EA was 1% (standard deviation 2.3%).

  The socio-cultural or self-contained concept of a race is not likely to be properly linked to human biology as a BGA, the genetic component of the race. Because it is subjective, inexact and sometimes inaccurate, the use of self-identified races for BGA inference is currently being implemented, but how and why human biology is human It is difficult to understand if it is related to anthropology. Still further, the inflexible segmentation of patients into pre-made racial groups is the implementation of a concept that is completely unsatisfactory, because many individuals divide their origin into multiple populations through a process of mixing. Because you can follow. An iterative, testable anthropological approach to defining BGAs, whether through direct correlation and / or through better research design, shows the relationship between BGAs and genetic disorders More sophisticated means can be provided, such as those that can be derived or through genetic mapping methods that rely on mixing processes such as MALD.

  As disclosed herein, 71 marker tests allowed the determination of BGA proportions and their confidence intervals. The test allowed determination of the relative proportion of BGA within an individual, thus distinguishing the BGA test from other tests previously used to infer ancestry from DNA. For the majority of BGA affiliations, more than 2200 tests were performed and no results were inconsistent with racial self-contained concepts. Previous ancestral testing was only accurate for the upper 90% range (see also Shriver et al., Supra, 1997; Fruitakis et al., Supra, 2003, which is incorporated herein by reference) . The improved performance observed in the BGA test is that CODIS and other STRs commonly used to infer ancestry from DNA are not selected for their δ values, but for their polymorphic complexity in the world population It may be because it was selected. For the BGA test disclosed herein, the entire genome was systematically scanned and the best AIM for this purpose was selected. Furthermore, most attempts to infer ancestry from DNA using STR or Alu sequences have attempted to classify or partition samples into a single “racial” group. For a wide range of individuals, such as a 50/50 mixture, such a method seems to produce a “wrong” answer as many times as a “correct” answer. In contrast, for BGA testing, ancestry is determined with respect to proportional affiliation and ameliorates this problem accordingly.

  The BGA test is distinguishable from other tests in that it uses SNPs that cover the majority of the chromosomes. Whole chromosome coverage using BGA testing offers substantial advantages over testing using CODIS STR, which only covers a small portion of the chromosome. Moreover, the BGA method appears to be the first to quantify the confidence limits for its answer. BGA tests, as illustrated, are widely heuristic and divide the world into four main anthropological groups that belong broadly along the continental line. Although geographic division respects the anthropological history of human migration, the use of the four groups is indeed a simplification of a very complex situation and can be considered optional. Furthermore, the problem of determining proportional affiliation has been simplified by calculating the most likely ternary (rather than quaternary) combination, because individuals with 4D BGA are considered rare And because it is more convenient in the sense of calculation. However, more complex tests can capture more of the real details of anthropological history, but even with a rough four-group test, the results can be used to build these categories and tests. It provides meaningful and historical content data if interpreted strictly for the parent sample.

  Choosing a particular parent group and choosing a simple partition of the world into a certain group will bring meaning to the coalescing time, thereby evaluating the reasoning given by the test. In fact, the difference between the answers given by a test based on 4 world population groups and a more complex test based on 25 would be one of anthropological time scales, but not "accuracy" . For example, most of the tested US-born Hispanics and most individuals claiming the tested American Indian heritage type showed a partial indigenous American mix in the Indo-European background. However, unlike the case with Hispanics, some individuals claiming American Indian heritage were classified as having a partial East Asian mix instead of an indigenous mix. Since the founders of Native Americans migrated from East Asia, perhaps in different eras of history, with different waves, the genetic distance between Native Americans and East Asians is between Native Americans and sub-Saharan Africa. Lower than between humans or Indo-Europeans (Cavalli-Sforza and Cavalli-Sforza, supra, 1995). Among pre-colonial North Americans, proportional affiliation to East Asians or indigenous Americans refers to individuals whose ancestors were part of the first wave, whereas those whose ancestors were part of the first wave Is expected to be different.

  The parent sample for the Native Americans used in this study was drawn from Southern Mexico, and the AIM allele frequency established for Native Americans obtained from this sample is that of early migration across the Bering Strait. It can be expected to represent more ancestors from the waves. Indigenous Americans from Latin America and South America, for example, from those from North America, such as Aruth Indian (and others), who probably belong more closely to their ancestors from late waves across the Bering Strait, It seems likely that they belong more closely to the early wave ancestors. East Asian affiliation for those individuals claiming affiliation with American Indians may be a by-product of the choice of South Mexico Nativos as a parent source and the use of only four groups of anthropological schemes But the answer is nevertheless a “wrong” answer in the scientific sense. Rather, it reports an affiliation with respect to the coalescence time scale defined by the parent sample source and an anthropologically meaningful way of dividing the world for this study.

  Different tests with more markers that can determine affiliation within the North American population from each other and from East Asians work on different coalescence time scales, and subdivide these individuals into “American Indian” or “North American It is likely to be classified as a “Native American” mix. Nonetheless, as demonstrated by the metrics incorporated into the test, the illustrated BGA test shows that a significant number of individuals of “Indigenous Americans” ancestry have more affiliations to East Asians than Native Americans. The fact that shows social and human history based on the concept of group affiliation is yet another example where social or human history is semantic and subjective and not necessarily accurate in a biologically meaningful way. It is possible. Even if the Aleutes seem to resemble East Asians in terms of physical characteristics, or even more than most Native Americans, and even if they are in temperate North America Even if they are indigenous to a geographical location as close to East Asia as they are, most people say that because their homes are east of the Bering Strait, North American Indians and By extension, is considered an indigenous American. A similar example is observed for certain other population groups, as discussed below, and measures of population affiliation using genetic markers and the return of racial identity by people from geographical and social boundaries. Illustrates the lack of communication between things that have been devised.

  It is interesting to compile the results to extract some meaningful anthropological and / or sociological knowledge, respecting the accreditation of the BGA test. Using the BGA test, 11 out of 201 African-Americans tested showed mostly Indo-European and small-African BGA, and most of the Puerto Ricans in the most African pedigrees They describe themselves as Hispanic and point out again the deficiencies in the current concept of race as a dichotomy based on artificial constructs. As observed in this study, Risch et al. (Supra, 2002) and Rosenberg et al. (Supra, 2002) found that most population affiliations were questioned when tested against methods that report BGAs that rely on genomic markers. It showed that it was reported quite accurately in the vote. This test of over 2,000 individuals shows that, for the most part, BGA affiliation can be accurately predicted from the self-declared race, the discrepancy between the two is not a significant event, and most ancestry affiliations The measurement shows that it is not a major problem with current self-reported methods. However, perhaps the most surprising result was the degree of mixing for each population examined. If the individual claimed mixing, it was almost always confirmed by BGA testing; in each case of expected sub-Saharan African and East Asian mixing, it was confirmed by BGA testing, and every Mexican Hispanic As expected, it showed either a majority or a minority of Native Americans.

  Slightly more than two-thirds of all “caucasians” examined showed a small East Asian or Native American mix, and virtually none of these individuals had a little in his or her questionnaire No significant pedigree mixing was reported. Some of this mixing seems to be a function of ethnicity. Not only did relatively uniform self-reported northern and eastern European heritage individuals more commonly exhibit East Asian BGAs, but Rosenberg et al. (Supra, 2002) supported this observation by "European Americans" or We have shown that there is a significant structure within the “European” population; specifically, they have shown that Russians generally exhibit small East Asian heritage. Such East Asian / Indo-European mixes may have become established many times in history, including the Mongolian invasion of Europe and the expansion of the Caucasian to the Scandinavian peninsula, Residents are of North Asian origin, exhibit Mongolian character, and share a common history and culture with East Asians (Cavalli-Sforza and Canvalli-Sforza, supra, 1995). Some of the observed Native American mixes are consistent with history; for example, the fact that the Filipinos show a wide range of Native American mixes has meant that the Spanish have conquered most of Latin America and until recently the Spanish territory It is not surprising that they exported native American slaves to these islands. The degree of indigenous Americans observed for a significant number of Filipinos is indeed high, perhaps the indigenous Americans are relatively common in this part of the world, and the genealogy for many Filipinos is high. Rather than the recent mix for individuals with a polarized BGA proportion, it reflects a domination with a large number of relatively low Native American mix individuals. The Native American mix commonly observed in “Caucasians” came from a mix of European and Native Americans in North America.

  Such as between Scandinavian / Russian IndoEuropean and East Asian, Sub-Saharan Africa / IndoEuropean in the US, Native American / East Asian mixed in the Philippines, or IndoEuropean / Indigenous American in the US In most cases of systematic mixing, the geographical proximity and / or historical mixing of related groups is well established by history. For example, African / East Asian mixes were rarely observed in these studies, and little history is known of the times when these two populations lived or mixed very close to each other . The type of mixing observed was also interesting when compared to the “racial” self-contained concept. For example, African-Americans are more mixed with Indo-European ancestry than Caucasians were mixed with African ancestry, and the difference in how Caucasians and Africans see their heritage Relieve and bring back the “one drop rule” recall. Assuming the degree of mixing observed, the concealment arising from the mixing process (because it is based on human constructs, a process that cannot be fully proven and quantified from our anthropological literature). There is a potential concern that other or potential BGA structures may cause rough (or fine) structural differences between groups of study samples. Such differences in structure would be expected to reduce the efficacy and power of large population-based study designs.

  For more than 2,200 blind subjects of polarized (low mix) and self-reported races, if the ancestry is largely consistent with the self-contained concept of affiliation, the accuracy of such minor mix proportions And how such accuracy can be measured if there is no BGA ultimate referee in existence (or see below, except for pedigree information). A problem occurs. Several experiments have been conducted to tackle this problem and, when considered together, indicate that fractional fractions are accurately measured. First, the fractional mixing percentage propagates along the pedigree in a manner consistent with the independent combination of genetic laws. By definition, a large unbiased error would make it appear that the fractional proportion was inconsistent in the pedigree relationship, i.e., assuming the parent proportion was correct, the independent combination law would be considered. If that is the case, the result will not occur.

  Second, the fractional mixing ratio, on average, is consistent with the self-contained concept of mixing. If there is a large, systematic and unbiased error in the BGA affiliation estimation, this error is more than the larger percentage percentage completeness because most individuals have a relatively polarized BGA affiliation. , Which will have a strong impact on the completeness of the smaller percentage, but if it is not so noticeable, the correlation of the smaller ratio / presence to the self-contained concept of mixing is probably very It will be weak. The fact was not observed this way. For example, if the error rate was around 20%, as many individuals as native Americans reporting small Hispanic ancestry showed small amounts of sub-Saharan African and East Asian ancestry. This result clearly demonstrates that this is not the case.

  Third, out of the approximately 2,200 blind-blown samples from North America, most individuals with a substantial (> 10%) sub-Saharan African mix, or vice versa, It was not observed at all. This non-observation is relevant because such individuals are extremely rare in North America from which validation samples were drawn. Where there was a large, unbiased and systematic error rate, the East Asian / African mixture would have been observed as frequently as the European-Asian mixture observed with high frequency.

  Fourth, a highly significant correlation was observed between a small Indo-European mix and skin melanin content in African Americans (see Example 1). If there was a large and unbiased error rate, such a correlation would not be likely to be observed.

  Fifth, as the true allele frequency for AIM related to affiliation decisions between the two groups was reduced by 20% for the AIM associated with each parental δ value for each given pair of groups When adjusted, the overall power of the BGA test has decreased with respect to determining the affiliation ratio between these two groups, but in essence, for the most part, and more importantly, The same result was produced for both submixture estimates. Parent sampling biases can thus cause inaccuracies in parental allele frequency estimates and δ values, but if the parent sample for this study consisted of about 100 individuals, it would definitely Less than 20%. Furthermore, errors in allelic frequency estimates would have to be obtained in the same direction for most of the AIMs associated with determining affiliation between group pairs for such errors that may exist. Nevertheless, this result indicates that the performance of the BGA test will be relatively unaffected even if such an error exists. In other words, this experiment demonstrated that the BGA test was relatively robust without parental sampling bias and allele frequency estimates.

  Sixth, the distortion of confidence contours along only the axis corresponding to the one associated with genotyping failure indicates a decent break in the interdigitated components of the examination (a subset of AIM). In other words, if the sample fits several templates for proportional fit measurements, the elements of these templates should be independent for the test to be valid and obtained The result was as follows.

  In light of the above six observations, it is difficult to imagine how significant systematic and unbiased errors can exist. Each observation by itself may not provide accuracy issues, but considering the results together means that there is little or no systematic and unbiased error in the disclosed BGA test. Provide sufficient evidence. Nevertheless, it can be argued that the errors in the inspection are not random and there is a linear style bias. However, the fifth observation (above) is a proof of this possibility. For example, the discovery of a small East Asian mix in an Indo-European background is more frequent than expected, with a “right” amount of allele frequency estimation error at the “right” number of markers in the “right” direction. Such a result can occur if is present (a situation that is unlikely but not impossible). However, since the AIM markers used in these studies are not mutually exclusive to a particular group of pairs, such errors are manifested in affiliation decisions for many pairs rather than just one And the first four observations (above) show that this is not the case. Also, such errors appear to require parent sampling from highly inbred groups. Many considerations have been made to avoid such sampling, but there are no tests that can be performed a priori, so the elements of the group were not always controlled. As such, a simulation was performed to estimate the contribution of linear error to the test results (fifth observation, above). The fact that the mixing results were relatively resistant to a substantial (20%) drop in δ values, the amount and quality of AIM used in this study was expected to be a reasonable level of sampling error Demonstrates that it had sufficient power so that it did not have a significant adverse effect on the quality of the results. Regarding the quality of the markers, considering the AIM selection process, this result seems reasonable because some of the best markers in the genome were used to measure BGA affiliation. With respect to the amount of markers, these results were very similar to those resulting from 30 marker tests (Shriver et al., Supra, 2003) with earlier results from 71 marker tests as disclosed herein. Is consistent with the observation that Thus, reducing the number of markers while retaining the average marker quality did not compromise the test. Furthermore, in another study, reducing the quality of the marker, not the quantity, produced the same result. Overall, these observations indicate that the BGA percentage produced by the test is accurate for a given confidence interval, and that the test is performed in a robust manner.

  The results demonstrate that BGA mixing is more common than previously thought. If true, it should be questioned whether a fine level of BGA mixing is linked to human biology, such as drug responsiveness or disease predisposition. Such “latent” structures can only be measured using molecular tests, unlike the general population structure, since questionnaires cannot be used. A fine level of collective structure has been recognized that exceeds the level of the coarser continent measured using questionnaires. For example, evidence of significant anecdotes suggests that redheads require 20% higher doses of many common anesthetics, and tend to hypertension and bleeding during anesthesia (Cohen, The Scientist 16:10, 2002). These complex physiological responses appear to be difficult to explain based on melanocortin-1 (MC1R) mutants previously associated with some of the changes in red hair color (Robbins et al. Cell 72: 827-834, 1993; Smith et al., J. Invest. Dermatol. 111: 119-122, 1998; Flanagan et al., Hum. Molec. Genet. 9: 2531-2537, 2000). If there are specific genetic variants that are likely to be responsible for these clinical phenotypes, and these variants correlate with elements of population structure or fine structure, they appear to be However, any study seeking to identify a linkage or LD between a marker and the associated phenotypically active locus appears to carry difficulties from the outset in the study design phase. The higher accuracy and objectivity than that provided by socio-cultural racial self-reporting may require the identification of structural elements that may interfere with the design of genetic experiments, so the illustrated BGA The test can be extended with AIM in addition to those disclosed herein to quantify the elements of the population structure associated with this particular problem.

Example 3
Application of BGA testing to genealogy This example demonstrates that BGA mixture estimation can be integrated with genealogical information obtained using traditional genealogical research methods.

  Genealogists are primarily concerned with geopolitical content (e.g. from which country an ancestor of a person comes) rather than in anthropological content (e.g. which type of population mix characterizes the family tree of the person). Collect relevant data), what was their religion, and data about their surname. There are two main sources for obtaining submixes in the results of one person: 1) recent territorial marriage mixed events; and 2) ancient affiliations to ethnic groups with the characteristics of systematic mixing.

  The consequences of extramarital marriage events are measured in the recent genealogical time (eg, in the last 250 years). For example, as shown in FIG. 11, a Chinese great-grandfather in an all-even uniform Indo-European pedigree gives rise to a mixed Indo-European / East Asian grandson. Individuals who are 100% East Asian (Chinese) are shaded (Fig. 11) and are subject to the mixed results for the lowest male (square) (short arrow) in the family tree. A person with one 100% East Asian great-grandfather and seven 100% Indo-European great-grandparents is expected to have a 12.5% East Asian mix. Due to the laws of genetic combinations, the expected level is actually a range around 12.5%, with values that can be up or down a few percent. The grandmother indicated by the long arrow is about 50% / 50% East Asian / Indo-European mixed, and her daughter, the mother of the subject, is 25% / 75% East Asian / Indian European mixed. Expected to be (Figure 11).

  Ancient affiliations (i.e., considered in terms of anthropological timeframes) are preserved in modern times due to the relative, geographically isolated, and tightly connected community structure (i.e. ethnicity) of family marriages. ing. For example, today's demographics are shaped not only by the migration that our ancestors made to establish new populations, but also by a mix between these populations throughout the world. The map below shows these migrations as measured from Y chromosome sequences generated over tens of thousands of years.

  Intermixing between groups, after each developing as a separate group, occurs many times throughout ancient and much more recent history, and is represented by arrows on the map representing migration patterns. For example, there is an extensive East Asian mix in Russians and Eastern Europeans (Rosenberg et al., Supra, 2003), and to some extent the Mongolian and Hun invasion may have contributed to this mix over time. Remains. There was a much more pronounced East Asian mix apparent for Native Americans (Rosenberg et al., Supra, 2003); much more arrows would have been needed and most of the mix events were unknown The arrow was not included on the map for this mixture. Nonetheless, people with a significant number of indigenous Americans or Russians in their pedigree are as east as individuals with a 100% Chinese grandmother and three other 100% IndoEuropean grandparents. The Asian mix was very good.

  Although not shown, the time scale was constructed to show when the most significant migrations occurred and corresponded to a very large family tree. The pedigree is for a single individual located at the bottom vertex of the triangle graph; if there are tens of thousands of ancestors for this person, it is large because it dates back to 60,000 years. The time scale for migration applies to large family trees as well. The family tree was the same as that shown on the family tree map (Figure 11), except that it was much larger and had no lines connecting the ancestors (the spots represented each ancestor, but there were so many It was not practical to show all the lines connecting the spots). The pool of spots represents "Russians" and for the purposes of this example was assumed to be a race born about 18,000 years ago. The additional spots represent East Asians, based on the assumption that the average Russians include a 10% East Asian mix. The third set of spots represented Russian pioneers; these pioneers were unknown but for the purposes of this example were assumed to be Eastern Europeans.

  In this example, the majority of the Russian ethnicity of this person came from the left side of the family tree, which can be assumed to be that part of the family tree that represents the father side of the family subject. As this example shows, if the average Russian contains a 10% East Asian mix and half of the family tree is predominantly Russian, then that person contains a 5% East Asian mix It is expected to be. Even if the person's grandmother, grandfather, or any other relative within the last 18,000 years is not a uniform East Asian, East Asian mixing is significant for this person. The way to visualize this on a family tree is to count all “East Asian” spots and divide them by the total number of spots in the family tree, which amounts to about 5%. Thus, relatively uniform East Asians represented about 5% of the total number of ancestors for this person. Of course, the family tree for some people includes multiple groups that are characterized by a small degree of this type of mixing. A family tree, such as the one illustrated, is polarized for a particular ethnicity and has an even distribution of each of the four BGA groups (sub-Saharan Africa, Native American, Indo-European, and East Asian) It is rare to see a family tree because, until recently and even now, people have tended to have others and children as well as themselves. As such, most pedigrees are highly polarized, as illustrated, rather than "random" of random affiliation.

  Pennsylvania Dutch provides another example of mixing by ancient affiliations, and it appears that there was significant East Asian content in the community before 1700 in the German ancestry. These ancestors established a community that occupies a place in the valley in the Upper Rhine basin, and later, inland farther away. Since these communities remained relatively isolated, the level of East Asians remained at about 20% level. This level of dilution requires external mixing with other Indo-European ethnic groups such as French or Sardinians who cannot detect East Asian mixing. In this case, the German ancestry had a substantial average East Asian mix, possibly due to a more heterogeneous sampling within the German population.

  Most genealogists are interested in the type of mixing found in the sources reached by the method of mixed marriage, which is more about the geopolitical affiliation of recent ancestors than the anthropological information about distant ancestry. Give information. This is because there is little document data about distant ancestors compared to modern people; the farther a person goes back in time, the greater the number of ancestors present, which is impossible If not, it makes the study difficult; and the contribution of distant ancestors to the genetic makeup of modern people on average is lower per relative than that of more recent ancestors. As such, genealogists tend to look for information such as what can happen with recent mixing. For example, if a person is trying to prove or disprove a rumor or legend of an American Indian ancestor, the 10% Native American mixed result was due to a recent mixed affiliation event with a mechanism for this mixing It would be very useful if it could be assured that it was not due to an ancient affiliation to ethnic groups with characteristics of systematic mixing. BGA testing does not allow a distinction between alien marriage versus ancient mixing, but the results of the test provide an important piece of genealogical puzzles.

  For some genealogists who rely on pedigrees, the evidence may strongly suggest that the mechanism of mixing is from recent events. As such, for a person whose family has documented evidence of an American Indian great-grandfather, a 10% Native American mixed result from a BGA test may indicate that the event was likely caused by recent mixing. In comparison, for a confirmed, uniform European ancestor, a 10% Native American mix indicates that the event is likely caused by an ancient mix.

  Negative results have a different meaning for genealogists than positive results. For example, if there is contextual but low quality data suggesting a purebred African great-grandfather and the BGA test reveals 100% Indo-European, the value of the rumor will be reduced (Taking into account the independent combination of genetic laws, which would make such a result possible but less likely if the data was indeed correct). However, if you suspect that a family member had a Chinese great grandfather, it is not possible to distinguish between a mixed alien marriage and an ancient one, so from the 20% East Asian mixed results Can not prove.

  It is important that genealogists use other knowledge to reconstruct the most likely source of mixed results. In fact, BGA mixed data serves as an independent clue to those trying to reconstruct a family, and when used with genealogical knowledge, the two combine to be more powerful than either alone To form evidence. As such, BGA testing is an anthropological source that communicated ancient or very old (including this 250-300 years in terms of genealogical timeframe) to us today, and events in the family tree over the last 200 years Providing BGA blending results in a manner that places equal emphasis on blending with family members by means of providing ancillary tools that can help shape custom systems for the pedigree community.

  A database of thousands of BGA profiles is built from people from various locations around the world, and you can query the database with profiles with or without a preselected error range. A list of places where this type of profile is commonly found can be provided, or a map of the world can be provided, for example by color, where the color can correspond to a mixed profile / range. Indicates high affiliation. In other words, given a BGA blending profile, the map can provide an indication of where the person's recent ancestors may have appeared. People with 10% East Asians and 90% Indo-Europeans have a high probability of ancestral origin from China (mixed marriages outside of tribe) or Russia (more ancient blends combined with ethnic homogeneity across the family tree) Indicates.

  Similarly, genealogists can provide similarly color-categorized maps derived from documentary studies that are based on geopolitical information rather than anthropological information. The two maps can then be overlaid and Bayesian statistical calculations can be made that combine the information from the BGA examination with the information in the document pedigree to provide the most likely estimate of a recent family tree.

  As an example, a source with a 90% Indo-European and 10% East Asian BGA and a Romanian / British / Spanish ancestry pedigree first queries the database with his or her BGA results. Maps are provided that are likely to be from East Asia (due to recent mixing), Russia and North / Eastern Europe (both by a number of distant ancestors from relatively isolated and mixed groups). The classification by color gives the probability of origin from a region based on the frequency with which a compatible BGA group is found in each region, which is the type of mixture, and the inventor who is the effect of sampling around the world. Depending on the characteristics of these databases, they can be quite complex. Second, he uses a map drafting tool that we can provide, and uses a separate map based on plausible Romanian / British / Spanish heritage documented from genealogical studies. (Or provided). From this map it is clear that the prospects for recent uniform East Asian ancestry are not high. Third, the program states that the most likely origin of 10% East Asian mixing is from Romanian ancestry (not British or Spanish and not by, for example, a Chinese grandfather). decide.

  This type of presentation allows one to know the most likely source of unexpected mixed results using previous knowledge gained through other means such as genealogical studies. This is valuable for genealogists who seek to elucidate the origin of the genetic makeup. Without this ancillary tool, those with 90% Indo-European and 10% East Asian mixing may have a test suggesting a recent Chinese or Japanese grandfather / great-grandfather, or a small East Asian mixing There seems to be no quick way to measure whether or not it suggested a particular ethnic affiliation commonly found.

Example 4
This example demonstrates that the potential population structure measured using AIM enables inferences about complex genetic traits such as iris color. ing.

  To measure whether and how common polymorphisms are associated with the natural distribution of iris colors, 851 individuals of mainly European descent were found in 13 chromogenic genes. It was examined in 335 SNP loci and 419 other SNPs that are known or thought to be distributed in the genome and inform about specific elements of the population structure. Numerous SNPs, haplotypes and diplotypes (haplotype diploid pairs) are significantly associated with OCA2, MYO5A, TYRP1, AIM, DCT and TYR genes, and iris color CYP1A2-15q22-ter, CYP1B1- Identified within the 2p21, CYP2C8-10q23, CYP2C9-10q24 and MAOA-Xp11.4 regions. Half of the relevant SNPs were located on chromosome 15, but others are consistent with results previously obtained from linkage analysis. Five additional genes (ASIP, MC1R, POMC and SILV) and one additional region (GSTT2-22q11.23) with haplotypes and / or diplotypes were identified but related to iris color It was not an individual's SNP allele (see also WO 02/097047). For most of the genes, the multilocus genotypic genotype sequence was more strongly associated with iris color than the haplotype or SNP allele. The diplotypes of these genes account for 15% of iris color variation. These results provide comprehensive candidate genetic studies for variable iris pigmentation and constitute a classification model useful for inferring iris color from DNA. The results further demonstrate that potential population structure can serve as a leverage tool for complex trait gene mapping when the genome is screened with the appropriate AIM.

  Iris pigmentation is a complex genetic trait that has long been of interest to geneticists, anthropologists and society in general, but is not yet fully understood. Eumelanin (brown pigment) is a light-absorbing polymer synthesized in differentiated melanocyte lysosomes called melanosomes. Within the melanosome, the tyrosinase (TYR) gene product catalyzes the rate-limiting hydroxylation of tyrosine to 3,4-dihydroxyphenylanine or DOPA, and the resulting product is oxidized to DOPA quinone for eumelanin synthesis Forming precursors. Although 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. Indeed, transmission genetics studies on pigmentation traits in humans and various model systems suggest that variable pigmentation is a function of multiple hereditary factors that appear to be quite complex in interaction. (See, eg, Akey et al., Supra, 2002; Box et al., J. Invest. Dermatol. 116, 224-229, 2001). For example, unlike human hair color (Sturm et al., Gene 277: 49-62, 2001), it appears that there is only a minor component in determining the color of a mammalian iris (Brauer and Chopra, Anthropol Anz. 36: 109-120, 1978), there is a minimal correlation between skin, hair and iris colors within or between individuals in a given population. In contrast, comparisons between populations show good agreement; populations with a black average iris color also tend to exhibit a black average skin tone and hair color. These observations indicate that genetic determinants for pigmentation in various tissues are distinct, and that these determinants are a common set of systematic and evolutionary forms that have shaped their distribution in the world population It suggests that it was easy to receive power.

  At the cellular level, the variable iris color in healthy humans is the result of differential deposition of melanin pigment granules within a certain number of interstitial melanocytes in the iris. Granule density appears to reach a genetically determined level by early childhood, with a minority of individuals showing a color change during later life, but usually throughout later life It remains unchanged. Pedigree studies have suggested that iris color variation is a function of two loci; one locus that causes depigmentation that does not affect skin or hair, and pigments in all tissues Another pleiotropic gene for the decrease in (Brues, Amer. J. Phys. Anthropol. 43: 387-391, 1975).

  Much of what is now known about pigmentation has been derived from molecular genetic studies of rare pigmentation defects in humans and model systems such as mice and Drosophila. For example, dissection of the ocular cutaneous albinism (OCA) trait in humans has shown that many pigmentation defects are due to damage in the TYR gene, resulting in a name as tyrosinase (TYR) negative OCAs (e.g. Oetting and King Hum. Mutat. 13: 99-115, 1999; see also the albinism database at the URL “cbc.umn.edu/tad/” on the World Wide Web (“www”). TYR catalyzes the rate-limiting step of melanin biosynthesis, and the degree to which the human iris is colored correlates well with the amplitude of the TYR message level. Nevertheless, the complexity of the OCA phenotype illustrated that TYR is not the only gene involved in iris pigmentation. Most TYR-negative OCA patients are completely depigmented, but the black iris albino mice (C44H) and their human IB ocular dermatoderma counterparts are pigmented in all tissues except the iris. Depletion is indicated (Schmidt and Beermann, Proc. Natl. Acad. Sci. USA 24; 91: 4756-4560, 1994). Numerous other TYR-positive OCA phenotype studies have been performed in addition to TYR, ocular dermatoderma 2 (OCA2; Durham-Pierre et al., Nature Genet. 7: 176-179, 1994; Durham-Pierre et al., Hum. Mutat. 7: 370-373, 1996; Gardner et al., Hum. Mutat. 7: 370-373, 1992), tyrosinase-like protein (TYRP1; Boissy et al., Amer. J. Hum. Genet. 58: 1145-1156, 1996) ), Melanocortin receptor (MC1R); Robbins et al., Supra, 1993; Smith et al., Supra, 1998; Flanagan et al., Supra, 2000) and adaptin 3B (AP3B; Ooi et al., EMBO J. 16: 4508-4518, 1997) The locus, as well as other genes (reviewed in Sturm et al., Supra, 2001) have been shown to be required for normal human iris pigmentation. Each of these genes is part of the major (TYR) human chromogenic pathway.

  In Drosophila melanogaster, iris pigmentation defects have been attributed to mutations in more than 85 loci contributing to various cellular processes in melanocytes (Ooi et al., Supra, 1997), but mouse studies have found about 14 Genes preferentially affect pigmentation in vertebrates (reviewed in Strum et al., Supra, 2001), and dissimilar regions of TYR and other OCA genes determine pigmentation in different tissues Suggested to be functionally distinct. Human chromogenic genes are for tyrosinase enzyme complex formation on the inner surface of melanosomes, hormonal and environmental regulation, melanoblast migration and differentiation, intracellular routing of new proteins to melanosomes, and melanosomes It develops into several biochemical pathways, including proper transport from the body of the cell to the dendritic arm towards the keratinocytes. Nevertheless, studies of human OCA mutants suggest that the number of phenotypically active chromogenic loci is small and manageable for genetic analysis.

  Studies in pigment mutants have revealed that a small subset of genes is primarily responsible for catastrophic pigmentation defects in mice and humans, but a common SNP in these genes is a natural occurrence in human iris colors. It remains unclear whether or how it contributes (or links) to fluctuations. The brown iris locus is located in the section containing the OCA2 and MYO5A genes (Eiberg and Mohr, Eur. J. Hum. Genet. 4: 237-241, 1996), and certain polymorphisms in the MC1R gene are relatively isolated Associated with red and blue iris colors (eg, Robbins et al., Supra, 1993; Flanagan et al., Supra, 2000; Valverde et al., Nature Genet. 11: 328-330, 1995; Schioth et al., Biochem. Biophys. Res. Comm. 260: 488-491, 1999). ASIP polymorphisms have been reported to be associated with both brown iris and hair color (Kanetsky et al., Amer. J. Hum. Genet. 70: 770-775, 2002). However, the penetrance of each of these alleles appears to be low, and generally only accounts for a very small amount of overall variation in iris color within the human population (Spritz Et al., Nature Genet. 11: 225-226, 1995). However, single gene studies did not provide a logically correct basis for understanding the complex genetic nature of human iris colors.

  Most human traits have complex genetic origins, and the whole is often larger than the sum of the parts, so genetic complexity--for example, the multifactorial nature of dominant and higher genetic variances and / or known components-- Innovative genomic science-based research design and analytical methods for screening genetic data with computers are needed. However, the first step is to define the complement of the locus that accounts for the variance in trait values at the sequence level, of which so doing in a marginal or osmotic sense, It seems the easiest to find. It is towards this goal that this study was conducted.

  A non-systematic, hypothetical-based genome screening approach has been applied to identify various SNPs, haplotypes, and diplotypes that are bounded (ie, independently) related to iris color variation. As disclosed in this example, a surprisingly large number of polymorphisms in many genes are associated with iris color, indicating that the genetic nature of iris color pigmentation is quite complex. Yes. The identified sequences provide the basis for classifier models for inferring iris colors from DNA, and some of these properties as markers of biogeographic ancestry map other complex trait gene mappings It has implications for study design.

Methods Specimen Collection Specimens for resequencing were obtained from the Coriell Institute in Camden, New Jersey. Samples for genotyping were self-reported Europeans of different ages, genders, hair, iris and skin darkness, and were collected using informed consent guidelines under IRB guidance . Donors check boxes about iris colors in blue, green, light brown, brown, black or unknown / unclear, each of which the iris color has changed over their life path We had the opportunity to identify whether or not each iris color was different. Individuals whose iris colors were ambiguous or changed over their course of life were excluded from the analysis.

  For 103 subjects, the iris color is also reported using a number between 1 and 11, with 1 being the darkest brown / black, 11 being the lightest blue identified using color marking is there. For these subjects, a digital photograph of the right iris was obtained, from which the subject stared into the box on one side and the camera on the other to standardize lighting conditions and distance. Samples were assigned to color groups by judgment. Comparing the two, 86 classifications matched. Of the 17 that were not, 6 were brown / light brown, 7 were green / light brown, and 4 were blue / green mismatched but brown / green, brown / blue or light brown / There were no terrible discrepancies like blue. While such errors are acceptable for identifying sequences that are borderlinely related to iris colors, by obtaining digitally quantified iris colors, the classification of iris colors is described herein. Confidence can increase with the use of the described sequences.

SNP Discovery Candidate SNPs were obtained from the NCBI: dsSNP database, which provides more candidate SNPs than are generally possible for genotypes. Human pigmentation and xenobiotic metabolism genes were examined and selected based on their gene identity rather than their chromosomal location. For some genes, the number of SNPs in the database was low, and / or some of the SNPs were strongly associated with iris color and would undertake deeper research. Of these genes, resequencing was performed; of the genes disclosed herein, 113 SNPs were CYP1A2 (7 gene regions, 5 amplicons, 10 SNPs CYP2C8 (9 gene regions, 8 amplicons, 15 SNPs found), CYP2C9 (9 gene regions, 8 amplicons, 24 SNP was found), OCA2 (16 gene regions, 15 amplicons, 40 SNPs were found), TYR (5 gene regions, 5 amplicons, 10 SNPs were found) and TYRP1 (7 gene regions, 6 amplicons, 14 SNPs were found; see Tables 9 and 10; WO 02/097047 Also see).

Table 9: Candidate genes tested for sequences associated with human iris pigmentation

  Resequencing for these genes was performed on 672 individuals (450 individuals from the Coriell Institutes DNA Polymorphism Discovery Resource, 96 additional European Americans, 96 African Americans and 10 Pacific Oceans. Multi-ethnic of islanders, 10 Japanese and 10 Chinese; these 672 represented a separate set of samples from those used in the association analysis described herein Proximal promoter from panel (average 700 bp upstream of transcription start site), each exon (average 1400 bp), 5 'and 3' ends of each intron (including intron-exon junction, average size about 100 bp) As well as by amplifying the 3 ′ UTR sequence (average 700 bp). PCR amplification was achieved using pfu TURBO polymerase according to the manufacturer's guidelines (Stratagene). The program is used to design primers that re-sequencing in a manner that favors homologous sequences in the genome to ensure that pseudogenes are not co-amplified or that sequences from within repeats are amplified. It was. A BLAST search confirmed the specificity of all primers used. The amplification product was subcloned into the pTOPO® sequencing vector (Invitrogen) and 96 insertion positive colonies were grown for plasmid DNA isolation (use of 670 individuals for the amplification step was selected. Reduced the likelihood of individuals contributing more than once to this subset of 96).

  Sequencing was performed on an ABI 3700 sequencer using PE Applied Biosystems BDT chemistry; the sequences were placed into a commercial relational database system (iFINCH, Geospiza; Seattle, WA). PHRED-certified sequences were imported into the CLUSTAL X alignment program and this output was used in a second program to identify quality-verified mismatches between sequences. Those sequences in which at least two examples of variants with a PHRED score of 24 or more were identified were selected and each of these SNPs discovered through resequencing was used for genotyping.

Genotyping:
For the majority of SNPs, the first round of PCR was performed on the samples using high fidelity DNA polymerase pfu TURBO polymerase and appropriate resequencing primers. A representative of the resulting PCR product was confirmed on an agarose gel, the first round PCR product was diluted and then used as a template for the second round of PCR. Since many of the genes queried were members of the gene family, two rounds were required; SNPs existed in the region of sequence homology, and the genotyping platform used a short (about 100 bp) amplicon. I needed it. For the remaining samples, only a single round of PCR was performed. Genotyping was performed on individual DNA specimens using a single base primer extension protocol and SNPstream ™ 25K / Ultra High Throughput (UHT) apparatus (Orchid Biosystems; Princeton, NJ). Required several quality controls; the two scientists independently observed an overall UHT signal intensity greater than 1,000 for> 95% and a clear signal difference between the means for each genotype class ( That is, the call was tested for pass / fail, requiring clear genotype cluster formation in 2-D space using UHT analysis software.

Statistical methods To test for deviations from independence in allelic states within and between loci, the MLD exact test was used (Zaykin et al., Supra, 1995). Haplotypes were inferred using haplotype reconstruction methods (Stephens et al., Supra, 2001). To measure the extent to which real iris color variations can be explained by various models, the R ² value is a phenotypic value of 1 for blue eye color and 2 for green eye color. The SNP, haplotype and multilocus genotype data were calculated by first assigning as 3 for the color eye color and 4 for the brown eye color. The BGA mixing ratio was measured as described within the context of software programs developed for this purpose. For R ² calculation, the following functions were used: Adj-R ² = 1-{(n / (nP)} (1-R ² ), where n is the model freedom and np is the error freedom An empirical Bayesian adjustment method for multiple results was used to correct for multiple tests (Steenland et al., Cancer Epidemiol. 9: 895-903, 2000, incorporated herein by reference. ).

ResultsTo identify SNP loci associated with variable human chromogenesis, chromogenic genes (AP3B1, ASIP, DCT, MC1R, OCA2, SILV, TYR, TYRP1, MYO5A, POMC, AIM, AP3D1 and RAB, Table 9 754 SNPs were genotyped, including 335 SNPs within and 419 other SNPs distributed throughout the genome. Alleles for these latter SNPs provided information about specific elements of population structure; their exceptionally high for BGA in Indo-Europeans, sub-Saharan Africans, indigenous Americans and East Asians 71 were selected from the screen of the human genome based on δ values (ie, exceptional AIM) (see Example 2; see SEQ ID NOs: 1-71; see also Shriver et al., supra, 2003), the rest being BGA It was found in or around genes for xenobiotic metabolism that tend to exhibit dramatic sequence variation as a function of. Genotypes for these 754 candidate SNPs were scored for 851 European-derived individuals (292 blue, 100 green, 186 light brown, and 273 brown) with self-reported iris colors It was done.

  Prior to screening these genotypes for association with iris color, 71 non-xenobiotic metabolism AIMs were used to measure the BGA mixing ratio for each sample, and between BGA mixing and iris color. Were examined for correlation. The test showed that each of the 851 Caucasian samples had mostly Indo-European BGA, and 58% of the samples had significant (> 4%) non-Indo-European BGA mixing, but low levels (33% (Less than) East Asian, Sub-Saharan African or Native American mix and iris color, and there was no correlation between high-level Native American mix and iris color; There was a weak association between higher-level (higher than 33% but lower than 50%) East Asian and sub-Saharan African blends and darker iris colors.

  It was uncertain from the beginning whether better success would be achieved by considering the iris color by four colors (blue, green, light brown and brown) or groups of colors. One way to classify colors is light = blue + green and dark = light brown + brown, and this classification seems to more clearly distinguish individuals with respect to detectable levels of eumelanin (brown pigment) It is. If the iris color data was self-reported, dividing the sample into brown and non-brown or blue and non-blue would have a significant association, especially for alleles associated with only one color. It can provide greater power to detect. In order to utilize each of these four methods, all were considered when screening SNPs for association; δ values, chi-squares and precision test p-values were a) all four colors, b) bright (blue) And green) versus darkness using dark (light brown and brown), c) blue versus brown, and d) brown versus non-brown (blue, green and light brown) classification. Significance level is fixed at 5%, 20 SNP alleles with specific iris color, 20 with iris color darkness, 19 with blue / brown color comparison, and 18 with Related using a brown / non-brown comparison. Overlap in these SNP sets was high but not complete; SNPs with significant p-values for associations using at least one of the four criteria were shown to be “boundary” related Yes.

  If multiple simultaneous hypotheses are tested in the set p-value, there is a possibility of an extended type I error. As such, a correction procedure was used to compensate for this risk (Steenland et al., Supra, 2002); most of the associations were significant after this correction. Most of the border-related SNPs are chromogenic genes-OCA2 (11 SNPs at the color level), TYRP1 (3 SNPs at the color level), MYO5A (2 SNPs at the color level), AIM (3 SNPs at the color level) and DCT (2 SNPs at the color level) − within, although some associations were found in CYP2C8 at 10q23, CYP2C9 at 10q24, 2p21 Found in non-pigmented genes such as MAOA at CYP1B1 and Xp11.3-these were also called border-associated SNPs. Although no significant SNP association was found within the chromogenic genes SILV, MC1R, ASIP, POMC, RAB or TYR, TYR had one SNP with p = 0.06. The most related of the border-related SNPs were derived from OCA2, TYRP1, and AIM genes in order of strength of association.

  Because most of the SNPs identified from this approach were located in distinct genes or chromosomal regions, all of the SNPs from each locus were grouped and the inferred haplotypes were matched with iris colors using contingency analysis. Tested for association. This more sophisticated analysis tested for grouped SNPs for all of the genes, not just those genes with borderline SNP associations. For each gene, haplotypes were inferred and contingency analysis was used to determine which haplotypes were statistically associated with iris color. From chi-square and modified residuals, 43 haplotypes at 16 different loci were associated with iris color, either positive (agonist) or negative (antagonist) (Table 10). . The strongest associations were observed for genes with border-associated SNPs; the majority of these genes were haplotypes and dipros that were positively (agonist) or negatively (antagonist) related to at least one iris color. Has a type (sometimes referred to as a polydentate genotype genotype or haplotype diploid pair) (Table 10). A few of the borderlessly related genes / regions that do not contain SNPs had haplotypes and diplotypes that were either positively and / or negatively related to iris color (one ASIP gene Haplotypes, MC1R-2 haplotypes, Table 10). In other words, those SNPs were only associated with iris color within a gene haplotype or diplotype relationship. For some, an association with iris color was only found within the diplotype relationship, but not at the SNP or haplotype level (ie SILV and GSTT2-22q11.23).

¹ 虹彩の色と関連していることが見出された遺伝子座内の最高オーダーの複雑性をもつ配列。少なくとも1つの有意に関連した配列をもつ各遺伝子座についての主要な配列(計数≧13)のすべてが示されている。遺伝子座についてのハプロタイプもディプロタイプも、関連していることが見出されなかった場合には、SNP対立遺伝子のみが示されている。ハプロタイプは遺伝子座について関連していることが見出されなかったが、ディプロタイプは関連していることが見出された場合には、ハプロタイプおよびディプロタイプの両方が示されている。
² アゴニスト色は、配列が正に関連している色を指す。アンタゴニスト色は、配列が負に関連している色を指す。カイ二乗P値が示されている。
³ 851個の本発明者らのサンプルにおいてハプロタイプが観察された回数。 Table 10: Common haplotypes and diplotypes for 16 iris color genes

^An array with the highest order of complexity within a locus found to be associated with ^one iris color. All of the major sequences (count ≧ 13) for each locus with at least one significantly related sequence are shown. If neither a haplotype nor a diplotype for the locus was found to be related, only the SNP allele is shown. Where haplotypes were not found to be related for a locus, but diplotypes were found to be related, both haplotypes and diplotypes are indicated.
² Agonist colors refer to colors that are positively related in sequence. Antagonist color refers to the color with which the sequence is negatively related. The chi-square P value is shown.
³ Number of times a haplotype was observed in 851 of our samples.

  At the haplotype level, each gene or region had a unique number and associated type. For example, OCA2, AIM, DCT, and TYRP1 included both haplotypes positively associated with blue iris and negatively associated with brown iris (OCA2 haplotype 1, 37, 38, 42, AIM haplotype 1, DCT haplotype 2, and TYRP1 haplotype 1, Table 10). Other genes such as AIM, OCA2 and TYRP1 were positively associated with brown, but included haplotypes associated with blue and negative (AIM haplotype 2, OCA2 haplotype 2, 4, 45, 47, TYRP haplotype 4, Table 10), while others, such as MYO5A, OCA2, TYRP1, and CYP2C8-10q23, included haplotypes that were positively associated with one color but not negatively associated with any other color ( MYO5A haplotype 5, haplotype 10, OCA2 haplotype 19, TYRP1 haplotype 3 and CYP2C8-10q23 haplotype 1, Table 10). The MC1R gene contained a haplotype that was only associated with green in our sample, and the POMC gene contained a single SNP with a genotype weakly associated with iris color (no significant haplotype or diplotype was found). Was not issued).

  Overall, the haplotype diversity associated with the brown iris was similar to that of the haplotype associated with the blue iris. The majority of haplotypes were more dramatically related to iris color in multi-racial samples because the large number of SNPs containing them are good AIMs, and darker iris colors and Related variants were enriched in those ancestry groups of the world with darker average iris colors. The majority of SNPs within a gene or region were in LD with others in that gene or region (D '<0.1); MC1R (1 pair), OCA2 (27 pairs), TYR (2 pairs) and TYRP1 Only 32 SNP pairs in the (2 pairs) gene were found to be in LD.

  These analyzes result in 16 SNPs associated with iris color at one or another level, whether related to a boundary or within a haplotype and / or diplotype relationship. This resulted in the identification of 61 SNPs in the gene / chromosomal region. The minor allele frequencies for the majority of these SNPs are relatively high (mean f minor allele = 0.22), and most of them were in Hardy-Weinberg equilibrium (HWE p> 0.05, 28/34, Table 10). Nine were not, and two of these were relatively infrequent and the evidence for an imbalance was a boundary (p value of almost 0.05). HWE deficiency is usually an indication of a poorly designed genotyping assay, and any of the remaining seven SNPs have a genotyping pattern that we have previously associated with such problems. Not shown (absence of one genotype class or heterozygous dominance). In fact, one of those with the strongest evidence of HWE deficiency was validated as a legitimate SNP through direct DNA sequencing. The distribution of SNP chromosomes that were significantly related in the boundary sense was independent of the distribution of SNPs actually examined, indicating that the association was not just the effect of SNP sampling.

  Chromosome 15q contains the majority of the SNPs (18/34) that were borderlinely related to iris color, and 14 of these chromosome 15 SNPs were found in two different genes, OCA2 and MYO5A. It was done. Chromosome 5p was all in the AIM gene with three border-associated SNPs, and chromosome 9p was all in the TYRP1 gene with five related SNPs. Multiple SNPs were identified on chromosome 10q; the CYP2C8-10p23.33 region had two SNPs and an adjacent region CYP2C9-10p24. All three markers were in LD in close contact with each other (p <0.001 for each possible pair). Multiple SNPs were also identified on chromosome 2; POMC SNPs located at 2p23 are borderlinely related, and SNPs from the CYP1B1-2p21 region are related within the context of the 2-SNP haplotype (Table 10), these SNPs were also present in LD (p <0.01). Finally, in addition to the OCA2 (15q11.2-q12) and MYO5A (15q21) sequences, a single SNP (15q22-ter) was also included on chromosome 15q, but each of these three loci There was no SNP in LD. Although SNPs for the MC1R (16q24), SILV (12q13), TYR (11q), MAOA-Xp11.4-11.3 and GSTT2-22q11.23 regions were also related at the haplotype level, these Was the only region of these chromosomes where an association was found.

  The obtained p-values indicated that the diplotype accounted for more iris color variation than the haplotype or individual SNPs. To test this, a corrected ANOVA analysis was performed on the data at each of these three levels. All 61 SNPs were considered, and their haplotypes (Table 10) and diplotypes (not shown) were also considered. After correcting for the number of variables, the diplotype accounted for 15% of the variation, while the haplotype accounted for 13% and the SNP accounted for 11% (Table 4). Of the 543 genotypes observed for 16 genes / regions, based on the chi-square modified residual, the 68 most strongly related genotypes accounted for 13% of the variation (line 4, table 11). ).

(Table 11) ANOVA-SNP and haplotype data

  A screen of 754 SNP loci identified 61 that were statistically associated with variable iris pigmentation at one or another level of complexity within the gene. The remaining SNPs had δ values and chi-square p-values that were not significant at any level of complexity within the gene. The diplotypes for these 61 alleles accounted for most of the iris color variance in the sample; at the level of SNP, the lowest amount was explained and the intragenic complexity to iris color determination Suggested the element of (ie dominance).

  Only about half of the 61 SNPs identified were independently associated with iris color; the rest were associated only in haplotype or diplotype relationships. Even at this level of complexity, sequences from non-single genes can be used to make reliable iris color inferences, and for iris color determination, aside from the complexity of the genes. Elements (ie superordinates) are also shown. Aside from the fact that many of the identified SNPs were significant after imposing a correction protocol for multiple testing, the five lines of evidence indicate that the identified SNPs are not falsely related. showed that. First, for all genes identified as border-related SNPs, multiple such SNPs were identified; that is, the distribution of SNPs among the various genes examined was not random . Second, some of the non-pigment gene SNPs are pigment genes such as CYP2C8 (10q24.1) and CYP2C9 (10q24) (two pigment genes that have not been tested directly-HPS1 (10q23.1-4 and HPS6). (Located proximal to (10q24.34)), the chromosome 2p SNP at the CYP1B1 locus (CYP1B1-2p21) is located proximal to POMC at 2p23 (and (In POMC SNP and LD) Third, roughly equal numbers of pigmented and non-pigmented gene SNPs were examined and had 34 border-associated SNPs, of which 28 ( 82%) were in chromogenic genes, and thus the distribution of SNPs among various gene types was also not random.Fourth, the association was generally of intragenic haplotypes The results were stronger for SNPs in the relationship and would not necessarily be obtained for falsely related SNPs (i.e. The results suggest that the gene sequences themselves are related, not just a single polymorphism within each gene.) Fifth, it includes multiple ancestral individuals When applied to a sample, linear and non-linear variables from these and other combined genes performed much better than when applied only to individuals of most European ancestry, or non-European or This improvement may not be surprising, as most individuals of small European descents exhibit low variability in iris color (on average, darker than European individuals). This result would not necessarily have been obtained if SNP was not truly associated with the color of the iris.

  The correction for multiple testing left most of the associations in the SNP level intact, but many associations did not pass the multiple testing, but were presented to avoid possible type II errors Sequences may be weakly related to iris color and possibly related within a multi-gene model for classification (ie superficiality). For these, downstream of SNP identification, such as from more sophisticated association tests, using various other criteria such as those described above, or for generalization of complex classification models It seems more sensible to eliminate false positives using the utility of.

  Mutations in the pigmentation gene are the main cause of ocular dermatoderma, so it is natural to expect that common mutations in their sequences explain some of the variance in natural iris color In fact, this result was observed. However, a number of its associations were with SNPs located in other types of genes (CYP2C8 at 10q23, CYP2C9 at 10q24, CYP1B1 at 2p21 and MAOA at Xp11.3). The inclusion of non-pigmented genes in this study was intentional; the screen is not limited to pigmented gene SNPs, but two types of AIMs—sub-Saharan, Indo-European, Native American and East Asian populations Included those selected from the genome based on δ values between allele frequencies, as well as those selected based on their location within the xenobiotic metabolism gene. Presumably, the xenobiotic metabolizing gene product is responsible for the detoxification of plant alkaloids and tannins present in land-specific foods, and that cocoons and genetic drift have shaped the geographical distribution of their sequences. Based in part on previous evidence showing that metabolic genes contain extraordinary concentrations of AIM, and that some of these AIMs are relevant to the measurement of “potential” population structure, the latter include. Such potential structures can correlate with iris colors to the extent that they allow accurate classification, even if they do not help elucidate biological mechanisms.

  a) Some of these SNPs are indicators of not only a rough or continental population structure, but also substructures and possibly even fine structures, b) the colors of the iris are those elements of structure within the Caucasian population And c) it was hypothesized that these markers could serve as a surrogate for phenotypically active loci for purposes of classification or trait value inference. The general concept is that genetic screening is only done strictly in the direction of identifying phenotypically active loci through linkage disequilibrium. However, if classification is the goal, population structure can help rather than identify phenotypically active loci if trait values correlate with structure and if markers for structure can be identified. For example, an iris color classification tool may be useful for forensic scientists to objectively and scientifically construct a partial physical profile from crime scene DNA. Currently, forensic investigators use surprisingly unscientific means to build physical profiles; rarely, witness reports are effective, and human reports are often subjective in certain situations and Untrusted. For forensic applications, investigators are less interested in phenotypic biological mechanisms than their ability to accurately infer trait values. Of course, identifying a marker in LD with a phenotypically active locus (or phenotypically active locus itself) provides a more accurate classification in addition to a better understanding of biological mechanisms, Searching for these hard to find loci in a heterogeneous population is impractical because LD only spans between a few Kb and requires costly genome-wide scans.

  The large number of SNPs identified herein as being associated with the iris color were located in the xenobiotic metabolism gene, indicating that the identified marker is iris color through correlation with potential population structure. Suggests that it is related to In other words, the non-pigmented gene marker is probably correlated with the phenotypically active locus for iris color, although not necessarily in LD. Through such a correlation, both markers and active loci are enriched in a particular branch of the Indo-European pedigree, even if they are not in each other and LD. These results based on such correlations are meaningful in classification relationships only with respect to the samples used. For example, AIMs selected based on their intercontinental delta values were not associated with iris color primarily in European individuals, but were strongly associated with iris color in more international samples. This is because AIM has a specific association with structural elements that correlate with iris color in this sample. In contrast, these same AIMs were not associated with iris color in most samples of individuals of European origin examined because there was little variation in the rough structure in this sample It is. Instead, in any (mostly) Caucasian or European-American sample, there is a substructure or fine structure (potential structure) due to variations in affiliation at the ethnic or other subgroup level. Apparently, if the structure correlates with the trait, only those SNPs specifically associated with measuring this potential structure would be required. The fact that no systematic structure was identified in a study sample of primarily European origin that was not related to the measured phenotype was irrelevant, and a certified AIM was used to reproduce this result for another trait. The use of is inevitable.

  Other interpretations of this result are possible, for example, that the association may have been observed through LD with what is not yet a limited chromogenic gene. In fact, CYP2C8 and CYP2C9 are located on chromosome 10 near the HPS1 and HPS2 chromogenic genes (not tested directly), and CYP1A2 is located on 15q22-ter on the same arm as OCA2 and MYO5A CYP1B1 is located at 2p21 near the POMC gene at 2p23, and MAOA is located on the same X chromosome arm (Xp11.4-11.3) as the OA1 chromogenic gene (directly) Have not been inspected automatically). The distance between these loci associated with iris color and the “adjacent” pigmentation genes is much larger than the average range of LD in the genome, even if these associations are through LD, It seems that the group structure needs to be quoted as an interpretation.

  LD is known to cross the megabase in recently mixed populations, and as few as 2,000 AIMs could be used to obtain whole-genome coverage in these populations, and Europe used in this study It is quite interesting that two-thirds of American-American samples had significant (4%) BGA mixing. European-Americans are not recognized as traditionally defined mixed groups (such as Hispanic or African-Americans), but the observed BGA mixes have a fine, potential level population structure. Can be tied together. The relevance of this result to LD and / or population structure is not clear, but if the result is based on LD rather than correlation, they are just for trait mapping in a population with a recent and extensive mix of AIM. It may be used to enhance population mixing, which also suggests that it can be used to enhance potential population structure in a similar manner. Thus, regardless of whether the outcome is by correlation or by LD, a large number of identified non-candidate gene associations are a cost-effective development of population structure measurements for pharmacogenomics and complex disease gene classifiers. It has a broader implication for.

  Linkage studies related specific chromogenic genes as being specifically related to the chromogenic phenotype, and the majority of chromogenic gene SNPs identified here are OCA2, MYO5A, TYRP1 and AIM. Clustered to such specific genes. Furthermore, certain aspects of the present case support previous literature. Most of the identified SNPs are on chromosome 15, which linkage analysis has identified as the main chromosome for the determination of “brownness” (Eiberg and Mohr, Eur. J. Hum. Genet. 4: 237-241 1996); suggesting that the candidate gene in the section containing this locus (BEY2) is most likely the OCA2 gene, although the MYO5A gene is also present in this section, and , As disclosed herein, associated with the color of the iris. OCA2 association was most significantly significant among the genes or regions examined, while MYO5A SNP was only weakly associated (but haplotypes and diplotypes were stronger). The MYO5A allele was not in LD with those of OCA2, and these results were obtained independently, and the results by Eiberg and Mohr may have been a reflection of the activity of two separate genes ( Eiberg and Mohr, supra, 1996).

  Two OCA2 code changes have been reported to be associated with darker iris color (Rebbeck et al., Cancer Epidemiol. Biomarkers Prev. 11 (8): 782-784, 2003). In addition, the “red hair / blue iris” SNP allele was previously identified (Valverde et al., Supra, 1995; Koppula et al., Supra, 1997), confirming that these sequences are associated with iris pigmentation, Nonetheless, the previously described association was with the blue iris and at the level of SNP, but in this study, the association was with the green iris and the haplotype and diplotype levels. It was only apparent in An association was also identified in the ASIP gene (Kanetsky et al., Supra, 2002), but in this study this genetic association was not at the level of SNP; one of the ASIP SNPs identified herein (Marker 861) is an 8818 GA SNP conversion described as being associated with the color of brown iris (Kanetsky et al., Supra, 2002), but in this study, the association is a light brown at the haplotype level. Met.

  The association between TYR haplotypes and iris colors is relatively weak and contradicts results obtained by others in the area of ocular dermatoderma who were unable to find a strong association in fewer samples. Absent. Although the results independently verified the findings for OCA2, ASIP and MC1R, they also found that several other pigmentation genes contain alleles associated with the natural distribution of iris color. Showing (TYRP1, AIM, MYO5A and DCT). As such, the results indicate that the previous work that linked the pigment-forming gene alleles to the iris color, which was taken up independently, represents only one stroke of a larger, more complex portrait. Show.

  Interestingly, the majority of SNPs identified herein are either silence phenotypes that are not translational regions, or are present in gene proximal promoters, introns or 3′UTRs. This result is not at all uncommon, but SNPs are at other phenotypically active loci and LDs, or that variability in message transcription and / or turnover is observed in the color of the human iris. It may indicate that this may be a reflection that a part can be explained. A large number of SNPs have been screened, but some of the genes contain a large number of candidate SNPs, and not all have been examined. For example, OCA2 has about 200 known candidate SNPs in NCBI dbSNP. As such, the OCA2 gene may still have more information on variable human iris pigmentation, and such information can be obtained using the methods disclosed herein.

Example 5
Use of AIM to predict drug responsiveness This example shows that, like many human genetic traits, AIM is chemically predictive and diagnostic because many drug response traits correlate with elements of population structure Demonstrates that it can be used to develop tests.

  The distribution of SNPs effective for genotyping by chromosomal arms is shown in FIG. In each of the approximately 400 SNPs examined, Caucasian individuals taking LipitorTM, whose responses were known for cholesterol (TC), low density lipoprotein (LDL), liver transaminase ASTSGOT and ALTGPT measurements ( 180 people) were genotyped. 150 Zocor ™ patients with known responses for TC and LDL changes, and 1,000 individuals with known hair and eye color were also genotyped. Those SNPs with significant delta values (δ> 0.20) in various trait classes were selected. For example, in about 70% of patients, Lipitor ™ caused a decrease in LDL, but in 30% of patients it had no effect. For any given SNP, the delta value (δ) is the difference in minor allele frequency among individuals with at least 20% reduction in LDL versus individuals with no such reduction in LDL. The δ value was measured for each SNP in FIG. 12 for each test (Zocor ™: LDL, TC, ASTSGOT, ALTGPT response; Lipitor ™: LDL, TC response). For eye color, δ values were measured by dark (light brown or brown) eyes versus light (blue or green) eyes. For hair color, δ values were measured by black or brown vs. blond.

  For the Lipitor ™ response, the number of SNPs with significant (δ> 0.20) values for each of the four endpoint measurements is shown in FIG. Those SNPs with significant delta values for the Zocor ™ response using LDL and TC endpoint measurements were then selected (FIG. 14). Next, those SNPs with significant delta values for iris color were selected (FIG. 15); and, similarly, for hair color (FIG. 16). The distribution of SNPs with good δ values was similar for each trait in graphs A to E, apart from specific elements of specificity. Specificity can be recognized by focusing on chromosomal arm 6p, which has many important SNPs for the TC (total cholesterol) response to LipitorTM (Figure 13), but the ZocorTM response There was nothing about (Figure 14). Chromosome 2 contains SNPs with good δ values for eye color (FIG. 16) but not hair color (FIG. 15). Chromosome 15 contains many markers that predict Lipitor ™ responsiveness, but not the Zocor ™ response. This specificity is likely the result of linkage disequilibrium with other loci that are deterministic for each of these traits, but nothing can manage the expression of the remaining traits; Type discovery is the traditional measurement goal of genetic mapping. Or it may be due to correlation with specific elements of the population structure.

  Similarities are evident in that chromosome 10 and 22 have a relatively high number of SNPs with good δ values for each of the four mechanistically unrelated traits, as does chromosome 1. Overall, the distribution of SNPs important for one trait does not differ from that for another. It is the similarity between profiles that is of interest here and that illustrates the value of the method.

  The elements that the four graphs have in common correlate with the number of SNP genotypes (Figure 12; Figures 13-16 are roughly similar to Figure 12). At the first barrier, this result is false in “importance” or significance for these SNP alleles and is simply a function of the number of SNPs genotyped for each chromosomal arm (i.e., chromosomes). The more SNPs that are genotyped, the more delta SNPs that will be found on that chromosome). However, the SNPs in FIG. 12 are not exactly any type of SNP; the majority of the effective SNPs in FIG. 12 are xenobiotic metabolism and chromogenic gene SNPs, almost all good AIMs.

  For iris colors, the majority of SNP associations remain significant (Chi p <0.05) after correction for multiple tests, thus indicating that SNP associations are not false. These experiments show that the distribution of related SNPs for each of the four traits is largely similar to each other, and that this distribution is similar to the distribution of effective SNPs that are mostly good AIMs. The majority of the SNPs measured in (1) are reporters of the population structure, and similar elements of the population structure respond to each of the two drugs (no matter how), plus hair And correlate with iris color pigmentation.

  It is noteworthy that these four traits appear to be mechanistically unrelated (at least for current knowledge), and the color of the hair or eyes depends on the response to the two randomly selected drugs. It is not intuitively clear as to how they can be related. However, the similarity in profiles for key SNPs for each of the traits suggests that each can be predicted to a significant extent with sequence knowledge about a common set of chromosomal markers. Since these markers are known to be excellent indicators of BGA, the results are more likely by each of the four unrelated traits measured by measuring the specific SNP measured in Figures 12-16. Rather, it shows that it can be predicted to some extent by measuring BGA.

  Simply measuring BGA as disclosed herein does not provide the predictive power for each of the four traits as measuring a particular AIM in the plot above. However, using the markers in FIG. 15 allows good classification accuracy for iris colors. These results show that the different AIMs in the above plot give information about different elements of the population structure or subgroup, and the majority of the SNPs in Figure 12 are good indicators of the rough population structure by continental BGA However, it has not yet been measured to what extent each provides information about other fine levels of structure, such as between Scandinavian and Mediterranean ancestry, or even within ethnic groups. Such “latent” structures are not possible prior to this disclosure to define sensitive elements of structures that are recognized in a large population of individuals of a common ethnicity in a biogeographically meaningful manner. Previously, it could not be defined in a reliable trust certification form. For example, most IndoEuropean redhead individuals are known to require 20% more anesthesia than other IndoEuropean (or other) patients (Cohen, supra, 2002). These redhead individuals also show a tendency to hypertension and bleeding under the influence of certain common anesthetics, thus presenting serious clinical problems of unknown etiology.

  Redhead individuals are common in the UK (Ireland and Britain), which also includes individuals that share more ancestors with each other than individuals in other parts of Europe. As such, they can be viewed as an uncertain structural branch that deviates from the human genealogy. Because only one gene is associated with red hair color (MCIR) and this gene has such a pleiotropic effect, its sequence varies in diverse and complex physiological responses such as those for multiple anesthetics It is difficult to imagine that it contributes to Furthermore, not everyone with red hair will contain a known red-haired MC1R variant, and usually will exhibit an abnormal anesthetic response. Rather, it is more likely that red hair color also correlates with elements of the population structure that correlate with anesthesia responses; that is, genes for red hair color and abnormal anesthetic responses are more common in these two traits Are specific or enriched to a particular branch of the human genealogy. Thus, the genes for red hair color and abnormal anesthetic responses are distributed as a function of population structure, and thus also have many other traits, as disclosed herein.

  The AIMs and methods disclosed herein are suitable for measuring various levels of structure, including rough continental structures as well as structures and potential structures related to ethnicity (e.g., Pacific Islands Almost 30 markers have been identified that can divide the system from other East Asians; that is, finer level structures than continental structures). AIM's informability is born across the process of evolutionary human development through founder effects, migration, bottlenecks, genetic drift and / or wrinkles, but these forces can affect hair color or anesthetic responses, There is no need to focus on the fact that the born AIM may correlate with these two types of traits. Just as in this case, there must be an AIM that informs about essentially different phenotypes in Europeans in the northwestern continent. No. The results demonstrating similarities in AIM with good delta values for Lipitor ™ response, Zocor ™ response, hair color and eye color (Figures 12-16) for each of these phenotypes Indicates the level of collective structure that provides information.

  It should be noted that the magnitude of association does not balance between traits. For example, the strongest AIM for iris color is not the strongest AIM for Lipitor ™ response, etc. Also, the direction of association is not necessarily the same in the trait; thus, an agonist association (positively associated) for the color of the blue iris is an agonist or antagonist (negatively associated with a particular Lipitor ™ response result). Any of the above).

  A randomly selected AIM that specifically distinguishes between Africans and East Asians is not required, but may contain information about a particular set of traits, because they are not required This is because it can be a marker of a specific element of the human population structure that correlates with traits (ie, a branch of the human genealogy where these traits are more common). Similarly, AIMs that distinguish between Indo-Europeans and Africans do not necessarily have the information they need to help predict anesthesia responses or redheads. Only those AIMs with alleles that distinguish between individuals of a particular fraction of humanity whose trait values are concentrated or displayed at a low rate are trait values, in this example anesthesia and red hair, or disclosed. As such, it possesses the information necessary to predict Lipitor ™, response to Zocor ™, hair color and eye color. A specific SNP can be a specific rough element of the population structure (European vs. sub-Saharan Africans), or substructure (Northern European vs. Indo-European of Mediterranean), or fine structure (Scottish vs. Irish vs. A good AIM for a British (or a Northern European with red hair vs. a Northern European with another hair color; or a Northern European who responds to a drug vs. another Northern European who does not respond to that drug). Some AIMs provide information about some level of population structure, others do not, and the majority of SNPs in the human genome do not carry any information at any level of the population structure (i.e. They are not AIM).

  The primary element of the disclosed method is that most human traits can be predicted through detailed measurements of AIM associated with population structure at various levels, provided that the trait correlates with that element of the structure. It is that. A secondary factor is that a classifier, or collection of SNP markers, and methods for predicting trait values from DNA can be constructed for most human traits through such recognition of population structure. As disclosed herein, such applications can be achieved through correlation as well as through a wide range of LDs found in certain mixed groups such as Hispanic or African Americans, For any sample of interest, regardless of race or ethnic background, the AIM used is a condition that it is appropriate for the elements of the population structure to which the trait is correlated. These results can be achieved because the method of the present invention mines the genome for good AIM, certifies their value as AIM, and accurately measures population structure against the background of the human phenotype.

  The trends observed in the studies represented by FIGS. 12-16 have been observed for many other traits, and SNP associations have such “penetration”, so they are corrected for multiple tests. (Steenland et al., Supra, 2000). As such, measuring AIM across the human genome predicts or infers the value of virtually any given human trait, even if the marker is not in linkage disequilibrium with the phenotypically active locus for the trait It is possible to know the elements of the collective structure, substructure or fine structure related to Because any human trait is more or less correlated with a particular element of the population structure, any human trait that is simple or complex, clinical, recreational, forensic or other value or not Correlation applies.

  As long as AIM is measured across the genome in an individual and it does not identify which (if any) correlates with the trait value, the trait segregates for population structure vs. substructure vs subordinate substructure vs. fine structure It is not possible to know a priori whether or not. Virtually any trait correlates with at least one element of structure, some with rough structure (as in human skin pigmentation) and some with substructure (South Asian Indo-European ( Indian) and Northern European Indo-Europeans (e.g. Irish), as in the case of iris, hair or skin pigmentation), and some with fine structure (in continental Europeans) Correlates, as in the case of red hair or anesthetic responses). It is not important to know which level of population structure a trait correlates to measure and find the AIM for the inference of that trait. Multiple general AIMs for statistical association with that trait It only needs to be measured and verified. As such, methods are provided for associating human gene sequences with traits so that they can be predicted or inferred. Such methods are valuable, for example, in the clinical and forensic field, because when predicting a common trait (drug response or disease predisposition that can cause disease), the general trait (drug response or disease This is because it is important to make an accurate inference of the predisposition), but it is not important that the biologically or mechanistically related sequences are measured.

Example 6
Algorithm for Maximum Likelihood Estimates A software program was written based on the algorithm of Hanis et al. (Supra, 1986) to determine the maximum likelihood estimate of an individual's BGA mixture using a multidentate AIM genotype. A flowchart illustrating an algorithm useful for determining proportional ancestry is provided in Table 12. An example of how the algorithm works is shown in Table 13, and the results of the ancestry percentage calculation are shown in Table 14.

The δ value is a representation of the marker ancestry information provision (Dean et al., 1994). For biallelic markers, the frequency difference (δ) is equal to p _x -p _y (equal to q _y -q _x ), where p _x and p _y are the frequencies of one allele in populations X and Y, q _x and q _y are the other frequencies. To test for deviations from independence in allelic status within and between loci, we used the MLD exact test (Zaykin et al., Supra, 1995). The collection of 71 AIMs used in Example 2 is cumulative within each of the six possible pairs of four-dimensional (sub-Saharan African, Native American, Indo-European and East Asian) problems. It was chosen to maximize the δ value and minimize the difference in accumulated δ values between each of the pairs.

  The algorithm reverses the population-specific allele frequency to obtain a likelihood estimate of proportional affiliation corresponding to the multilocus genotype using three groups at once; mainly for computational convenience and also 4 Three groups were used because dimensional mixing is likely to be relatively rare. For example, the likelihood of 100% Indian Europeans, 0% Native Americans, 0% East Asians is calculated, then 99% Indian Europeans, 1% Native Americans, 0% East Asians are calculated next And so on until all possible Indo-European, Native American and East Asian percentages are considered, and then the process is followed by all possible Indo-European, Native American and African percentages , And repeated for all possible Native American, African and East Asian proportions. The likelihood of the maximum value is selected as the maximum likelihood estimate (MLE). When plotting a single MLE on a triangle plot, the space where the likelihood is within 2x, 5x and 10x MLE is delimited (see Figure 3); multiple MLEs are a single triangle plot In general, these intervals are not plotted.

Table 12: Algorithm for ancestor calculation-flow chart

I. 目隠しサンプルについて最良の3つの集団を選び取る
II. 最大尤度値をもつ割合を得る
1. 最良の3つの集団を選び取る：
アルゴリズム：
すべての集団について
{
すべてのSNPについて
{
集団合計<-集団合計+期待遺伝子型頻度。
}
}
最大値をもつ3つの集団を選び取る。
段階1：
すべての集団について
段階2：
SNP1は不均一の遺伝子型をもつ；対立遺伝子はGおよびTである。
SNP1についての期待遺伝子型=log(2*P(G,1)*P(T,1))；
SNP2は均一の遺伝子型をもつ
SNP2についての期待遺伝子型=log(P(T,1)*P(T,1))；
尤度値集団1=SNP1についての期待GT+SNP2についての期待TT
すべての集団について段階2を繰り返す
それらの4つの集団値から最良の3つの尤度値を選び取る。
それらの選択された3つの集団推定割合について。
1. 以下から始まる
λ₁=0、λ₂=0およびλ₃=1 0+0+1=1
2. 尤度値を計算する：
SNP1について期待遺伝子型を推定する
SNP1は不均一の遺伝子型をもつ；対立遺伝子はGおよびTである。
サンプルからの推定対立遺伝子頻度：
対立遺伝子1推定頻度(A1EF)=λ₁.p(G,1)+λ₂.p(G,2)+λ₃.p(G,3)
p(G,1)−集団1におけるG対立遺伝子頻度
p(G,2)−集団2におけるG対立遺伝子頻度
p(G,3)−集団3におけるG対立遺伝子頻度
混合している割合λ₁、λ₂およびλ₃は未知のパラメーターとして処理される。
対立遺伝子2推定頻度(A2EF)=λ₁.p(T,1)+λ₂.p(T,2)+λ₃.p(T,3)
p(T,1)−集団1におけるT対立遺伝子頻度
p(T,2)−集団2におけるT対立遺伝子頻度
p(T,3)−集団3におけるT対立遺伝子頻度
パラメーターの尤度は、ハーディ-ワインベルグの法則の仮定の下、新しい観察においてそれぞれ観察された遺伝子型についての確率に掛けることにより得られる。
SNP1は不均一の遺伝子型をもつため
SNP1についての期待遺伝子型=log(2*A1EF*A2EF)；
SNP2について期待遺伝子型を推定する
サンプルからの推定対立遺伝子頻度：
対立遺伝子1推定頻度(A1EF)=λ₁.p(T,1)+λ₂.p(T,2)+λ₃.p(T,3)
p(T,1)−集団1におけるT対立遺伝子頻度
p(T,2)−集団2におけるT対立遺伝子頻度
p(T,3)−集団3におけるT対立遺伝子頻度
混合している割合λ₁、λ₂およびλ₃は未知のパラメーターとして処理される。
SNP2は均一の遺伝子型をもつため
SNP2についての期待遺伝子型=log(A1EF*A1EF)；
尤度値
すべてのSNPのすべての期待遺伝子型を加えることにより尤度値を計算する
尤度=SNP1についての期待遺伝子型+SNP2についての期待遺伝子型；
異なる未知のパラメーターを用いることにより尤度値を計算する。(段階2を繰り返す)
3. 最大尤度値および対応する未知のパラメーターを得る。
それらの未知のパラメーターは、割合以外にない。 Table 13: Example of proportional ancestor determination using algorithm

I. Pick the best three populations for blindfold samples
II. Get the fraction with the maximum likelihood value
1. Pick the best three groups:
algorithm:
About all groups
{
About all SNPs
{
Population total <-Population total + Expected genotype frequency.
}
}
Select the three groups with the maximum values.
Stage 1:
Stage 2 for all groups:
SNP1 has a heterogeneous genotype; alleles are G and T.
Expected genotype for SNP1 = log (2 * P (G, 1) * P (T, 1));
SNP2 has a uniform genotype
Expected genotype for SNP2 = log (P (T, 1) * P (T, 1));
Likelihood value group 1 = Expectation TT for SNP1 Expected TT for SNP2
Pick the best three likelihood values from those four population values that repeat step 2 for all populations.
About those three selected population estimates.
1. Starting with λ ₁ = 0, λ ₂ = 0 and λ ₃ = 1 0 + 0 + 1 = 1
2. Calculate the likelihood value:
Estimate expected genotype for SNP1
SNP1 has a heterogeneous genotype; alleles are G and T.
Estimated allele frequency from sample:
Allele 1 estimated frequency (A1EF) = λ ₁ .p (G, 1) + λ ₂ .p (G, 2) + λ ₃ .p (G, 3)
p (G, 1) -G allele frequency in population 1
p (G, 2) -G allele frequency in population 2
The proportions λ ₁ , λ ₂ and λ ₃ mixing in p (G, 3) -population ₃ are treated as unknown parameters.
Allele 2 estimated frequency (A2EF) = λ ₁ .p (T, 1) + λ ₂ .p (T, 2) + λ ₃ .p (T, 3)
p (T, 1)-T allele frequency in population 1
p (T, 2)-T allele frequency in population 2
The likelihood of the T allele frequency parameter in p (T, 3) -population 3 is obtained by multiplying the probability for each genotype observed in each new observation under the assumption of Hardy-Weinberg law.
SNP1 has a heterogeneous genotype
Expected genotype for SNP1 = log (2 * A1EF * A2EF);
Estimated allele frequencies from samples that estimate expected genotypes for SNP2:
Allele 1 Estimated Frequency (A1EF) = λ ₁ .p (T, 1) + λ ₂ .p (T, 2) + λ ₃ .p (T, 3)
p (T, 1)-T allele frequency in population 1
p (T, 2)-T allele frequency in population 2
The proportions λ ₁ , λ ₂ and λ ₃ mixing in p (T, 3) -population ₃ are treated as unknown parameters.
SNP2 has a uniform genotype
Expected genotype for SNP2 = log (A1EF * A1EF);
Likelihood value is calculated by adding all expected genotypes of all SNPs. Likelihood value = Likely expected genotype for SNP1 + Expected genotype for SNP2;
A likelihood value is calculated by using different unknown parameters. (Repeat step 2)
3. Obtain the maximum likelihood value and the corresponding unknown parameter.
There are no other unknown parameters.

例

III. 目隠しサンプルについて最良の3つの集団を選び取る
IV. 最大尤度値をもつ割合を得る
a. 3つの最良の集団を選び取る
アフリカ人(目隠しサンプル100%アフリカ人を仮定する)：
SNP1対立遺伝子：G、T
アフリカ人における「G」対立遺伝子頻度 P(G)=0.8
アフリカ人における「T」対立遺伝子頻度 P(T)=0.2
SNP1についての期待遺伝子型値=log(2*P(G)*P(T))
=log(2*0.8*0.2)
=-0.4948
SNP2対立遺伝子：T、T
アフリカ人における「T」対立遺伝子頻度 P(T)=0.7
SNP2についての期待遺伝子型値=log(P(T)*P(T))
=log(0.7*0.7)
=-0.3098
SNP3対立遺伝子：C、T
アフリカ人における「C」対立遺伝子頻度 P(C)=0.9999
アフリカ人における「T」対立遺伝子頻度 P(T)=0.0001
SNP3についての期待遺伝子型値=log(2*P(C)*P(T))
=log(2*0.9999*0.0001)
=-3.6990
アフリカ人についての尤度=-0.4948-0.3098-3.6990
=-4.5036
ヨーロッパ人(目隠しサンプル100%ヨーロッパ人を仮定する)：
SNP1対立遺伝子：G、T
ヨーロッパ人における「G」対立遺伝子頻度 P(G)=0.9
ヨーロッパ人における「T」対立遺伝子頻度 P(T)=0.1
SNP1についての期待遺伝子型値=log(2*P(G)*P(T))
=log(2*0.9*0.1)
=-0.7447
SNP2対立遺伝子：T、T
ヨーロッパ人における「T」対立遺伝子頻度 P(T)=0.7
SNP2についての期待遺伝子型値=log(P(T)*P(T))
=log(0.7*0.7)
=-0.3098
SNP3対立遺伝子：C、T
ヨーロッパ人における「C」対立遺伝子頻度 P(C)=0.8
ヨーロッパ人における「T」対立遺伝子頻度 P(T)=0.2
SNP3についての期待遺伝子型値=log(2*P(C)*P(T))
=log(2*0.8*0.2)
=-0.4948
2. ヨーロッパ人についての尤度=-0.7447-0.3098-0.4948
=-1.5493
先住アメリカ人(目隠しサンプル100%NAを仮定する)：
SNP1対立遺伝子：G、T
NAにおける「G」対立遺伝子頻度 P(G)=0.6
NAにおける「T」対立遺伝子頻度 P(T)=0.4
SNP1についての期待遺伝子型値=log(2*P(G)*P(T))
=log(2*0.6*0.4)
=-0.3187
SNP2対立遺伝子：T、T
NAにおける「T」対立遺伝子頻度 P(T)=0.5
SNP2についての期待遺伝子型値=log(P(T)*P(T))
=log(0.5*0.5)
=-0.6020
SNP3対立遺伝子：C、T
NAにおける「C」対立遺伝子頻度 P(C)=0.7
NAにおける「T」対立遺伝子頻度 P(T)=0.3
SNP3についての期待遺伝子型値=log(2*P(C)*P(T))
=log(2*0.7*0.3)
=-0.3767
3. 先住アメリカ人についての尤度=-0.3187-0.6020-0.3767
=-1.2974
中東(目隠しサンプル100%MEを仮定する)：
SNP1対立遺伝子：G、T
MEにおける「G」対立遺伝子頻度 P(G)=0.7
MEにおける「T」対立遺伝子頻度 P(T)=0.3
SNP1についての期待遺伝子型値=log(2*P(G)*P(T))
=log(2*0.7*0.3)
=-0.3767
SNP2対立遺伝子：T、T
MEにおける「T」対立遺伝子頻度 P(T)=0.9
SNP2についての期待遺伝子型値=log(P(T)*P(T))
=log(0.9*0.9)
=-0.0915
SNP3対立遺伝子：C、T
MEにおける「C」対立遺伝子頻度 P(C)=0.9
MEにおける「T」対立遺伝子頻度 P(T)=0.1
SNP3についての期待遺伝子型値=log(2*P(C)*P(T))
=log(2*0.9*0.1)
=-0.7447
4. 中東についての尤度=-0.3767-0.0915-0.7447
=-1.2129
アフリカ人についての尤度=-4.5036
5. ヨーロッパ人についての尤度=-1.5493
6. 先住アメリカ人についての尤度=-1.2974
7. 中東についての尤度=-1.2129
この場合、本発明者らは、アフリカ人を落とし、割合について他の3つを考える。
最大尤度値
そこで、いくつかの値を未知のパラメーターへ与えることを始める
I=ヨーロッパ人 J=先住アメリカ人 K=中東
常に、I+j+k=1
I=0；j=0；k=1
{
SNP1対立遺伝子：G、T
ヨーロッパ人における「G」対立遺伝子頻度 P(G,1)=0.9
NAにおける「G」対立遺伝子頻度 P(G,2)=0.6
MEにおける「G」対立遺伝子頻度 P(G,3)=0.7
対立遺伝子1推定頻度(A1EF)=I*P(G,1)+J*P(G,2)+K*P(G,3)
=0*0.9+0*0.6+1*0.7
=0.7
ヨーロッパ人における「T」対立遺伝子頻度 P(T,1)=0.1
NAにおける「T」対立遺伝子頻度 P(T,2)=0.4
MEにおける「T」対立遺伝子頻度 P(T,3)=0.3
対立遺伝子2推定頻度(A1EF)=I*P(T,1)+J*P(T,2)+K*P(T,3)
=0*0.1+0*0.4+1*0.3
=0.3
SNP1についての期待遺伝子型値=log(2*A1EF*A2EF)
=log(2*0.7*0.3)
=-0.3767
SNP2対立遺伝子：T、T
ヨーロッパ人における「T」対立遺伝子頻度 P(T,1)=0.7
NAにおける「T」対立遺伝子頻度 P(T,2)=0.5
MEにおける「T」対立遺伝子頻度 P(T,3)=0.9
対立遺伝子1推定頻度(A1EF)=I*P(T,1)+J*P(T,2)+K*P(T,3)
=0*0.7+0*0.5+1*0.9
=0.9
SNP2についての期待遺伝子型値(EGV2)=log(A1EF*A2EF)
=log(0.9*0.9)
=-0.0915
SNP3対立遺伝子：C、T
ヨーロッパ人における「C」対立遺伝子頻度 P(C,1)=0.8
NAにおける「C」対立遺伝子頻度 P(C,2)=0.7
MEにおける「C」対立遺伝子頻度 P(C,3)=0.9
対立遺伝子1推定頻度(A1EF)=I*P(C,1)+J*P(C,2)+K*P(C,3)
=0*0.8+0*0.7+1*0.9
=0.9
ヨーロッパ人における「T」対立遺伝子頻度 P(T,1)=0.2
NAにおける「T」対立遺伝子頻度 P(T,2)=0.3
MEにおける「T」対立遺伝子頻度 P(T,3)=0.1
対立遺伝子2推定頻度(A1EF)=I*P(T,1)+J*P(T,2)+K*P(T,3)
=0*0.2+0*0.3+1*0.1
=0.1
SNP3についての期待遺伝子型値(EGV3)=log(2*A1EF*A2EF)
=log(2*0.9*0.1)
=-0.7447
i. 未知のパラメーターについての尤度値
=EGV1+EGV2+EGV3
=-0.3767-0.0915-0.7447
=-1.2129
ヨーロッパ人=0；NA=0；中東=1について；尤度値は-1.2129である
}
すべての可能な組み合わせについて上のループを繰り返す
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,
0.1,0.2,0.7,
0.1,0.3,0.6,
0.1,0.4,0.5,
など
最大尤度値を得て、対応する割合が祖先割合である。 (Table 14) Calculation of ancestry ratio
Ancestor frequency table

Example

III. Pick the best three populations for blindfold samples
IV. Get the ratio with the maximum likelihood value
a. Select the three best groups
African (assuming 100% African with blindfold sample):
SNP1 alleles: G, T
“G” allele frequency in Africans P (G) = 0.8
'T' allele frequency in Africans P (T) = 0.2
Expected genotype value for SNP1 = log (2 * P (G) * P (T))
= log (2 * 0.8 * 0.2)
= -0.4948
SNP2 alleles: T, T
'T' allele frequency in Africans 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 Africans P (C) = 0.9999
'T' allele frequency in Africans P (T) = 0.0001
Expected genotype value for SNP3 = log (2 * P (C) * P (T))
= log (2 * 0.9999 * 0.0001)
= -3.6990
Likelihood for Africans = -0.4948-0.3098-3.6990
= -4.5036
European (assuming 100% European blindfold sample):
SNP1 alleles: G, T
"G" allele frequency in Europeans P (G) = 0.9
“T” allele frequency in Europeans P (T) = 0.1
Expected genotype value for SNP1 = log (2 * P (G) * P (T))
= log (2 * 0.9 * 0.1)
= -0.7447
SNP2 alleles: T, T
"T" allele frequency in Europeans 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 Europeans P (C) = 0.8
"T" allele frequency in Europeans P (T) = 0.2
Expected genotype value for SNP3 = log (2 * P (C) * P (T))
= log (2 * 0.8 * 0.2)
= -0.4948
2. Likelihood for Europeans = -0.7447-0.3098-0.4948
= -1.5493
Native Americans (assuming a blindfold sample of 100% NA):
SNP1 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. Likelihood for Native Americans = -0.3187-0.6020-0.3767
= -1.2974
Middle East (assuming 100% ME blindfold sample):
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 SNP1 = log (2 * P (G) * 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 SNP2 = 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 SNP3 = log (2 * P (C) * P (T))
= log (2 * 0.9 * 0.1)
= -0.7447
4. Likelihood for the Middle East = -0.3767-0.0915-0.7447
= -1.2129
Likelihood for Africans = -4.5036
5. Likelihood for Europeans = -1.5493
6. Likelihood for Native Americans = -1.2974
7. Likelihood for the Middle East = -1.2129
In this case, we drop Africans and consider the other three in terms of proportion.
Maximum likelihood value So start giving some values to unknown parameters
I = European J = Indigenous American K = Middle East Always, I + j + k = 1
I = 0; j = 0; k = 1
{
SNP1 alleles: G, T
"G" allele frequency in Europeans P (G, 1) = 0.9
“G” allele frequency in NA P (G, 2) = 0.6
“G” allele frequency in ME P (G, 3) = 0.7
Allele 1 estimated frequency (A1EF) = 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 Europeans P (T, 1) = 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 (A1EF) = I * P (T, 1) + J * P (T, 2) + K * P (T, 3)
= 0 * 0.1 + 0 * 0.4 + 1 * 0.3
= 0.3
Expected genotype value for SNP1 = log (2 * A1EF * A2EF)
= log (2 * 0.7 * 0.3)
= -0.3767
SNP2 alleles: T, T
"T" allele frequency in Europeans 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 (A1EF) = I * P (T, 1) + J * P (T, 2) + K * P (T, 3)
= 0 * 0.7 + 0 * 0.5 + 1 * 0.9
= 0.9
Expected genotype value for SNP2 (EGV2) = log (A1EF * A2EF)
= log (0.9 * 0.9)
= -0.0915
SNP3 alleles: C, T
“C” allele frequency in Europeans 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 (A1EF) = I * P (C, 1) + J * P (C, 2) + K * P (C, 3)
= 0 * 0.8 + 0 * 0.7 + 1 * 0.9
= 0.9
"T" allele frequency in Europeans P (T, 1) = 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 (A1EF) = 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 * A1EF * A2EF)
= log (2 * 0.9 * 0.1)
= -0.7447
i. Likelihood values for unknown parameters
= EGV1 + EGV2 + EGV3
= -0.3767-0.0915-0.7447
= -1.2129
European = 0; NA = 0; Middle East = 1; likelihood value is -1.2129
}
Repeat the 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,
0.1,0.2,0.7,
0.1,0.3,0.6,
0.1,0.4,0.5,
The maximum likelihood value is obtained, and the corresponding ratio is the ancestor ratio.

  Although the invention has been described with reference to the above examples, 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 claims.

Claims (82)

A method for inferring an individual's traits with a predetermined confidence level, including the following steps:
a) contacting a sample containing a nucleic acid molecule of a test individual with a hybridizing oligonucleotide, wherein the hybridizing oligonucleotide exhibits at least about 10 ancestral informational markers (AIM) exhibiting a population structure correlated with the trait The nucleotide occurrence of a single nucleotide polymorphism (SNP) in this panel can be detected, and the contacting step is carried out under conditions suitable for detecting the nucleotide occurrence of the AIM of the test individual by the hybridizing oligonucleotide. Stage, and
b) identifying a population structure that correlates with the occurrence of AIM nucleotides in a test individual with a predetermined confidence level, wherein the population structure correlates with a trait, thereby inferring the individual's trait with a predetermined confidence level; Stage.   The method of claim 1, wherein the panel comprises at least about 20 AIMs.   2. The method of claim 1, wherein the trait comprises biogeographic ancestry (BGA).   4. The method of claim 3, wherein the panel comprises AIM shown as SEQ ID NOs: 1-331.   4. The method of claim 3, wherein the panel comprises AIM shown as SEQ ID NOs: 1-71. 4. The method of claim 3, wherein the panel comprises AIM indicated as:
SEQ ID NOs: 7, 21, 23, 27, 45, 54, 59, 63 and 72-152;
SEQ ID NOs: 3, 8, 9, 11, 12, 33, 40, 59, 63 and 153-239;
SEQ ID NOs: 1, 8, 11, 21, 24, 40, 172 and 240-331; or combinations thereof.   2. The method of claim 1, wherein at least one AIM of the panel is not linked to a gene associated with the trait.   4. The method of claim 3, wherein the BGA comprises a percentage of sub-Saharan African, Native American, Indo-European, or East Asian ancestry groups, or a combination of the ancestry groups.   9. The method of claim 8, wherein the BGA comprises a proportion of at least 3 ancestral groups.   BGA of at least sub-Saharan African and Indo-European ancestry groups; Native American and Indo-European ancestry groups; East Asian and Indigenous American ancestry groups; or Indo-European and East Asian ancestry groups 4. The method of claim 3, comprising a percentage.   4. The method of claim 3, wherein the BGA comprises a proportion of at least Native American, East Asian and IndoEuropean ancestry groups; or Sub-Saharan African, Native American and IndoEuropean ancestry groups.   2. The method of claim 1, wherein the trait comprises an individual's responsiveness to a drug.   13. The method of claim 12, wherein the drug is a cancer chemotherapeutic agent.   13. The method of claim 12, wherein the drug is a statin.   2. 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 according to claim 16, wherein the disease is cancer, diabetes or hypertension.   18. The method of claim 17, wherein the cancer is prostate cancer.   16. The method of claim 15, wherein the disease is a neurological disorder.   20. The method of claim 19, wherein the method is schizophrenia or Parkinson's disease.   16. The method of claim 15, wherein the disease is alcoholism.   2. The method of claim 1, wherein the trait comprises a chromogenic trait.   23. The method of claim 22, wherein the pigmentation trait comprises eye color, skin color, hair color, or a combination thereof.   The method of claim 1, further comprising identifying a subpopulation structure of a population structure that correlates with the occurrence of AIM nucleotides in a test individual with a predetermined confidence level, wherein the subpopulation structure is correlated with a trait. Method.   The hybridizing oligonucleotide comprises an oligonucleotide primer, and the method further comprises contacting the sample with a polymerase under conditions suitable for production of the primer extension product, wherein measuring the nucleotide occurrence of the SNP is a primer extension product. 2. The method of claim 1, comprising detecting the presence of.   The hybridizing oligonucleotide comprises an oligonucleotide primer, and the method further comprises contacting the sample with a polymerase under conditions suitable for generation of a primer extension product, and measuring the nucleotide occurrence of the SNP comprises: 2. The method of claim 1, comprising measuring the nucleotide sequence of the primer extension product at a position corresponding to the position.   The hybridizing oligonucleotide comprises an amplification primer pair, and the method further comprises contacting the sample with a polymerase under conditions suitable for production of the amplification product, wherein determining the nucleotide appearance of the SNP is the presence of the amplification product 2. The method of claim 1, comprising detecting.   The hybridizing oligonucleotide comprises an amplification primer pair, and the method further comprises contacting the sample with a polymerase under conditions suitable for the generation of an amplification product, wherein measuring the nucleotide occurrence of the SNP comprises measuring the position of the SNP 2. The method of claim 1, comprising measuring the nucleotide sequence of the amplification product at a position corresponding to.   The method of claim 1, wherein the method is performed in a high throughput format.   The method of claim 1, wherein the method is performed in a multiplexed format. A method for estimating a proportional ancestor of at least two ancestry groups of individuals with a predetermined confidence level, including the following steps:
a) contacting a sample containing a nucleic acid molecule of a test individual with a hybridizing oligonucleotide, wherein the hybridizing oligonucleotide represents at least about a biogeographic ancestry (BGA) for each ancestral group being examined. A single nucleotide polymorphism (SNP) nucleotide occurrence of a panel of 10 ancestral informative markers (AIM) can be detected, and the contacting step can detect the nucleotide appearance of the test individual's AIM by the hybridizing oligonucleotide. Under conditions suitable for detection; and
b) identifying a population structure that correlates with the nucleotide occurrence of each AIM of the ancestry group being examined, with a predetermined confidence level, wherein the population structure indicates a proportional ancestry, thereby determining an individual's proportional ancestry Estimating with a confidence level of.   32. The method of claim 31, wherein the proportional ancestry comprises a proportion of sub-Saharan African ancestry groups, Native American ancestry groups, Indo-European ancestry groups, East Asian ancestry groups, or combinations thereof.   Proportional ancestry is sub-Saharan African and Indo-European ancestry groups; Native American and Indo-European ancestry groups; East Asian and Indigenous American ancestry groups; or Indo-European and East Asian ancestry groups 32. The method of claim 31, comprising a proportion of:   Proportional ancestry of Native American, East Asian and Indo-European ancestry groups; Sub-Saharan African, Native American and Indo-European ancestry groups; or Sub-Saharan African, Native American and East Asian ancestry groups 32. The method of claim 31, comprising a proportion of ancestral groups.   32. The method of claim 31, wherein the panel for at least one of the ancestral groups comprises the AIM set forth in SEQ ID NOs: 1-331.   32. The method of claim 31, wherein the panel for at least one of the ancestral groups comprises an AIM shown as SEQ ID NOs: 1-71. 32. The method of claim 31, wherein the panel for at least one of the ancestry group comprises an AIM indicated as:
SEQ ID NOs: 7, 21, 23, 27, 45, 54, 59, 63 and 72-152;
SEQ ID NOs: 3, 8, 9, 11, 12, 33, 40, 59, 63 and 153-239;
SEQ ID NOs: 1, 8, 11, 21, 24, 40, 172 and 240-331; or combinations thereof.   32. The method of claim 31, wherein at least one AIM of the panel is not linked to a gene associated with the trait. 32. The method of claim 31, wherein the step of identifying a population structure wherein the proportional ancestry comprises a proportion of three ancestry groups and correlates with the AIM nucleotide occurrence of the test individual comprises the following steps:
Determining likelihood for affiliations to sub-Saharan African ancestry groups, Native American ancestry groups, IndoEuropean ancestry groups and East Asian ancestry groups;
Then selecting the three ancestor groups with the largest likelihood values;
Determining the likelihood of all possible proportional affiliations among the three ancestral groups with the largest likelihood values, thereby correlating the population structure or proportional affiliation with the AIM nucleotide occurrence of the test individual As well as identifying only one proportional combination of maximum likelihood, thereby estimating the proportional ancestry of the individual. 32. The method of claim 31, wherein the step of identifying a population structure in which the proportional ancestry comprises a proportion of three ancestry groups and correlates with the nucleotide occurrence of AIM comprises the following steps:
Making 6 binary comparisons, including determining the likelihood of affiliation with each other in each group;
Then selecting the three ancestor groups with the largest likelihood values;
Determining the likelihood of all possible proportional affiliations among the three ancestral groups with the largest likelihood values, thereby correlating the population structure or proportionality with the AIM nucleotide occurrence of the test individual Identifying the global affiliation; and identifying only one proportional combination of maximum likelihood, thereby estimating the individual's proportional ancestry.   Identifying a population structure whose proportional ancestry contains a proportion of three ancestry groups and correlates with the AIM nucleotide appearance of the tested individual performs three ternary comparisons between the groups, the largest likelihood value Determining the likelihood of all possible proportional affiliations among the three ancestral groups having, thereby identifying a population structure or proportional affiliation that correlates with the nucleotide occurrence of the AIM of the test individual, 32. The method of claim 31, comprising identifying; and identifying only one proportional combination of maximum likelihoods, thereby estimating the proportional ancestry of the individual.   Identifying a population structure where the proportional ancestry contains a proportion of the four ancestry groups and correlates with the nucleotide occurrence of the AIM of the test individual is six binary comparisons, three ternary comparisons, or 1 Performing four quaternary comparisons; determining the likelihood of all possible proportional affiliations among the four ancestral groups with the highest likelihood values, thereby determining the nucleotide occurrence of the AIM of the individual being tested 32. The method of claim 31, comprising: identifying correlated population structure or proportional affiliation; and identifying only one proportional combination of maximum likelihoods, thereby estimating a proportional ancestry of the individual .   Creating a graphical representation of a comparison of the three ancestry groups, each graph representation including a triangle, each ancestor group independently represented by a vertex of a triangle, and the maximum likelihood of proportional affiliation for the individual 41. The method of claim 40, further comprising the step of: the degree value includes a point in the triangle.   44. The method of claim 43, wherein the graphical representation further comprises a confidence contour indicating a confidence level associated with estimating proportional ancestry.   Further comprising identifying a subpopulation structure of the population structure that correlates with the occurrence of AIM nucleotides in the test individual with a predetermined confidence level, wherein the subpopulation structure correlates with the ethnicity of the test individual. Item 31. The method according to Item 31. 46. The method of claim 45, wherein identifying the subpopulation structure comprises the following steps:
Identifying the chromosome of the test individual that contains the AIM that represents the ancestral group of the test subject's proportional ancestry;
Contacting the sample containing the nucleic acid molecule of the test individual with a second hybridizing oligonucleotide, wherein the second hybridizing oligonucleotide can detect the nucleotide appearance of the SNP in the second panel of AIM. And a second panel of AIMs are present on the chromosome of the test individual, including the AIM representing the tester's ancestry group; and a subpopulation structure that correlates with the nucleotide appearance of the second panel of AIMs A stage, wherein the sub-group indicates the ethnicity of the ancestral group of the test individual.   48. The method of claim 45, wherein the ancestry group is Indo-European and the ethnicity includes Northern European or Mediterranean.   32. The method of claim 31, further comprising: creating a global ancestry map, wherein the location of a population having a proportional ancestry corresponding to the proportional ancestry of the test individual is indicated on the ancestor map. . 49. The method of claim 48, further comprising the following steps:
a) overlaying an ancestor map with a phylogenetic map, where the phylogenetic map indicates the location of a population with geopolitical relevance with respect to the examined individual; and
b) statistically combining ancestral and phylogenetic information to obtain the most probable estimate of the test individual's pedigree.   The method of claim 31, wherein identifying a population structure that correlates with the nucleotide occurrence of AIM comprises comparing the AIM nucleotide occurrence of the test individual to a known proportional ancestor corresponding to the nucleotide occurrence of AIM indicative of BGA. Method.   51. The method of claim 50, wherein the database includes known proportional ancestry corresponding to the nucleotide occurrence of AIM indicative of BGA.   52. The method of claim 51, wherein the comparing is performed using a computer.   51. The method of claim 50, wherein each of the known proportional ancestry corresponding to the nucleotide occurrence of AIM indicative of BGA is further comprised of a photograph of the person for 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 including digital photographs is included in the database.   56. The method of claim 55, wherein the digital information in the database is associated with a known proportional ancestor corresponding to the nucleotide occurrence of AIM indicative of the person's BGA in the photograph.   52. The method of claim 51, further comprising identifying a photograph of a person having a proportional ancestry that corresponds to the proportional ancestry of the examined individual.   AIM showing the steps of identifying a photo, scanning a database containing multiple files, each file containing digital information corresponding to a digital photo of a person with a known proportional ancestor, and the BGA of the individual being examined 58. The method of claim 57, comprising identifying at least one photograph of a person having an AIM nucleotide occurrence indicative of a BGA corresponding to the nucleotide occurrence.   A product comprising at least one photograph of a person with a known proportional ancestry corresponding to a population structure containing nucleotide occurrences of AIM representing biogeographic ancestry (BGA).   60. The article of claim 59 contained in a file.   60. The plurality of files comprising the product of claim 59, wherein the plurality of files comprises at least one photograph of a person with a known proportional ancestry corresponding to a population structure that includes a nucleotide occurrence of AIM indicative of BGA.   61. The file of claim 60, wherein the plurality of photographs includes a plurality of photographs, including photographs of persons with known proportional ancestry corresponding to a population structure that includes nucleotide occurrences of AIM indicative of BGA.   64. The file of claim 62, wherein the photos of the plurality include photos of different people having the same known proportional ancestry.   64. The file of claim 62, wherein the photos of the plurality include photos of different people with different known proportional ancestry.   60. The product of claim 59, wherein the at least one photo comprises a digital photo.   68. The product of claim 65, wherein the digital photograph includes digital information.   68. The product of claim 66, wherein the digital information is included in the database.   68. The product of claim 65, comprising a plurality of digital photographs.   66. The plurality of products of claim 65, comprising at least two digital photographs.   70. The plurality of claim 69, wherein the digital photograph includes digital information.   72. The plurality of claim 70, wherein the digital information is included in the database.   A kit comprising a plurality of hybridizing oligonucleotides comprising at least 15 contiguous nucleotides of at least 5 polynucleotides shown in SEQ ID NOs: 1-331, or polynucleotides complementary thereto.   75. The kit of claim 72, wherein the hybridizing oligonucleotide comprises at least 15 contiguous nucleotides of at least 5 polynucleotides shown in SEQ ID NOs: 1-71, or polynucleotides complementary thereto.   75. The kit of claim 72, wherein the hybridizing oligonucleotide of the plurality comprises at least one nucleotide corresponding to a polymorphic position of the polynucleotide or a polynucleotide complementary thereto.   The hybridizing oligonucleotide of the plurality is a SEQ ID NO: 1-34, 36-49, 52-55, or 57-98, 100-105, 107-162, 164-331, or a complementary polynucleotide thereof. 75. The kit of claim 74, comprising nucleotide position 50 of the polynucleotide shown in any of the nucleotides.   73. The kit of claim 72, wherein the hybridizing oligonucleotide of the plurality comprises at least one probe, at least one primer, or a combination thereof.   77. The kit of claim 76, comprising at least one amplification primer.   77. The kit of claim 76, comprising at least one amplification primer pair comprising a forward primer and a reverse primer.   79. The kit of claim 78, further comprising a reagent for performing an amplification reaction using at least one amplification primer pair.   73. The kit of claim 72, wherein the ancestral information providing marker (AIM) further comprises at least one AIM corresponding to the hybridizing oligonucleotide of the plurality.   75. The kit of claim 72, further comprising a detectable label that can be bound to or incorporated into at least one hybridizing oligonucleotide of the plurality.   73. The kit of claim 72, wherein the plurality of hybridizing oligonucleotides are detectably labeled.

Images