JP2001514906A - Methods and compositions for detecting or quantifying nucleic acid species - Google Patents

Abstract

(57) Abstract The present invention provides an array of probes immobilized on a substrate and a method for detecting a target nucleic acid species using a large number of labeled probes. The present invention also relates to oligonucleotide probes attached to discrete particles that can be classified into multiple sets based on physical properties. Different probes are bound to each set of discrete particles, and the identity of the probe is determined by identifying the discrete particles from their physical properties. The invention further employs agents that destabilize (reduce binding energy) binding of complementary polynucleotide strands or increase the stability of binding (increase binding energy) between complementary polynucleotide strands. About the method. The drawing is a schematic view of an apparatus for mass-producing a probe array.

Description

DETAILED DESCRIPTION OF THE INVENTION

This application is a continuation-in-part of US patent application Ser. No. 08 / 912,885, filed Aug. 15, 1997, and US patent application Ser. / Part of 947,779, and 1
Partial continuation of U.S. patent application Ser.No. 08 / 892,503 filed on Jul. 14, 997, and part continuation of U.S. patent application Ser. This is a continuation-in-part of US Patent Application No. 08 / 784,747, filed on the 16th.

FIELD OF THE INVENTION [0002] The present invention relates generally to methods and apparatus for nucleic acid analysis, and more particularly to methods and apparatus for nucleic acid analysis.

BACKGROUND The speed of sequencing four nucleic acids in a nucleic acid sample is a major technical obstacle to further advances in molecular biology, medicine, and biotechnology. Nucleic acid sequencing, which separates nucleic acid molecules in gels, has been used since 1978. Another proven method for nucleic acid sequencing is sequencing by hybridization (SBH).

[0004] The traditional method of determining the nucleotide sequence (ie, the order of A, G, C and T nucleotides in a sample) is to prepare a mixture of nucleic acid fragments having random ends and different labels, Degradation at specific nucleotides or done by the dideoxy chain ends of the replicating strand. The resulting nucleic acid fragments ranging from 1 to 500 bp are then separated on a gel to create a ladder of bands where adjacent samples differ by one nucleotide in length.

[0005] The array-based approach of SBH does not require single base digestion in the separation, degradation, synthesis or imaging of nucleic acid molecules. Hybridization that identifies mismatches in short oligonucleotides of K bases in length can be used to determine the list of constituent K-mer oligonucleotides for a target nucleic acid. The sequence for the target nucleic acid can be constructed with oligonucleotides that are scored as uniquely overlapping.

[0006] Several approaches are available to achieve sequencing by hybridization. In a process called SBH format 1, nucleic acid samples are arrayed and labeled probes are hybridized to the samples. Replica membranes with the same set of sample nucleic acids can be used for parallel scoring of several probes and / or probes can be multiplexed. Nucleic acid samples can be arrayed and hybridized on a nylon membrane or other suitable support. Each membrane array can be reused many times. Format 1 is particularly efficient for batch processing of large numbers of samples.

In the SBH format 2, probes are arrayed at positions on the substrate corresponding to the respective sequences, and a labeled nucleic acid sample fragment is hybridized to the arrayed probes. In this case, sequence information for a fragment can be determined by a simultaneous hybridization reaction with all array sequence probes. The same oligonucleotide array can be reused for sequencing other nucleic acid fragments. The array can be made by spotting or by in situ synthesis of probes.

[0008] In SBH format 3, two sets of probes are used. In one embodiment, one set can be in the form of an array of probes with known locations, and the other labeled set can be stored in a multi-well plate. In this case, there is no need to label the target nucleic acid. A target nucleic acid and one or more labeled probes are added to the arrayed probe set. If one bound probe and one labeled probe both hybridize in close proximity to the target nucleic acid, they ligate covalently to produce a detection sequence equal to the sum of the ligated probe lengths. I do. The process allows long nucleic acid fragments, such as the entire bacterial genome, to be sequenced without the nucleic acid subcloning into small pieces.

In the present invention, SBH has applications in the efficient identification and sequencing of one or more nucleic acid samples. The method has many applications in nucleic acid diagnostics, forensics, and gene mapping. It can also be used to identify mutations associated with genetic disorders and other traits, assess biodiversity, and generate many other types of data that depend on nucleic acid sequence.

SUMMARY OF THE INVENTION [0010] The present invention is a method for detecting a target nucleic acid species, comprising providing an array of probes immobilized on a substrate and a number of labeled probes, wherein each labeled probe is a target nucleic acid species. One having a first nucleic acid sequence that is complementary to a portion is selected, and wherein the nucleic acid sequence of at least one probe immobilized on a substrate is complementary to a second portion of the target nucleic acid sequence and The portion is adjacent to the first portion; the target nucleic acid is added to the array under conditions suitable for hybridizing the probe sequence with the complementary sequence; the labeled probes are introduced into the array; and immobilized on a substrate. Allowing the probe to hybridize to the target nucleic acid; hybridizing the labeled probe to the target nucleic acid; immobilizing the labeled probe to an adjacent hybridized probe in the array; It provides the above method comprises detecting the labeled probe fixed to the probe in the array by. According to a preferred method of the present invention, the array of probes immobilized on a substrate comprises a universal set of probes. In another preferred aspect of the invention, at least two probes immobilized on the substrate define overlapping sequences of the target nucleic acid sequence, and more preferably at least two labeled probes define overlapping sequences of the target nucleic acid sequence. I do. And according to yet another aspect of the invention,
A method for detecting a target nucleic acid of known sequence, comprising contacting a nucleic acid sample with a set of immobilized oligonucleotide probes bound to a solid substrate under hybridization conditions, wherein the immobilized probes comprise different portions of the target nucleic acid sequence. Contacting a target nucleic acid with a set of labeled oligonucleotide probes in solution under hybridization conditions, wherein the labeled probe is a different portion of the target nucleic acid sequence adjacent to an immobilized probe. Is capable of specific hybridization with the probe; covalently connecting the immobilized probe (eg, with a ligase) to a labeled probe immediately adjacent to the immobilized probe on the target sequence;
Removing the unligated labeled probe; detecting the presence of the target nucleic acid by detecting the presence of the labeled probe bound to an immobilized probe. The present invention also relates to a method of determining the expression of a member of a set of partially or completely sequenced genes in a cell type, tissue or tissue mixture, comprising: Defining a pair of labeled probes; hybridizing an unlabeled nucleic acid sample and corresponding labeled probes to an array of one or more immobilized probes; sharing between adjacent hybridized labeled and immobilized probes. Forming a bond; removing unligated probes; and determining the presence of the sequenced gene by detecting a labeled probe bound to a pre-specified position in the array. provide. In a preferred embodiment of this aspect of the invention, the target nucleic acid will identify the presence of an infectious agent.

Further, the present invention relates to a nylon membrane (a plurality of sub-arrays of oligonucleotide probes on the nylon membrane, the sub-array comprising a plurality of individual spots, wherein each spot comprises a plurality of oligonucleotides having the same sequence). And an array of oligonucleotide probes comprising a plurality of hydrophobic barriers (located between the subarrays on the nylon membrane, thereby preventing cross-contamination between adjacent subarrays).

Further, the present invention provides a method for determining the sequence of a repetitive sequence having a first end and a second end of a target nucleic acid, comprising: (a) a plurality of spacer oligonucleotides of various lengths; Providing a spacer oligonucleotide comprising said repeat sequence; (b) providing a first oligonucleotide known to be adjacent to the first end of the repeat sequence; Providing a plurality of second oligonucleotides adjacent to the second end of the sequence, the plurality of second oligonucleotides being labeled; (d) the first and plurality of second oligonucleotides, and the plurality of spacer oligonucleotides; Hybridizing one of the lignonucleotides to the target nucleic acid; (e) ligating the hybridized oligonucleotide; (f) ligating the oligonucleotide. Separating the tide from the unligated oligonucleotide; and (g) detecting a label in the ligated oligonucleotide.

[0013] Still further, the invention is a method for sequencing a branch point sequence having a first end and a second end within a target nucleic acid, comprising: Providing a first oligonucleotide that is complementary to the moiety, wherein the first oligonucleotide extends at least one nucleotide from a first end of the branch point sequence;
(b) providing a plurality of second oligonucleotides that are labeled and are complementary to the second portion of the branch point sequence, wherein the plurality of second oligonucleotides are at least one from the second end of the branch point sequence; (C) the portion of the second oligonucleotide extending by nucleotides and extending from the second end of the branch point sequence, comprising a sequence that is complementary to a plurality of sequences that result from the branch point sequence; Hybridizing the first oligonucleotide and one of the plurality of second oligonucleotides with the target DNA; (d) ligating the hybridized oligonucleotide; (e) not ligating the ligated oligonucleotide. Separating from the oligonucleotide; and (f) detecting the label of the ligated oligonucleotide.

[0014] Further, the present invention provides a method for confirming a sequence using a probe predicted to be negative for a target nucleic acid. In that case, the sequence of the target will
By hybridizing with "negative" probes to ensure that these probes do not form a perfect match with the target nucleic acid.

[0015] Still further, the present invention provides a method for analyzing nucleic acids using oligonucleotide probes complexed with different labels so that the probes can be multiplexed in a hybridization reaction without losing sequence information ( That is, different probes have different labels so that hybridization of the different probes to the target can be distinguished). In a preferred embodiment, the label is a radioisotope, or a fluorescent molecule, or an enzyme, or an electrophoretic mass label. In a further preferred embodiment, differently labeled oligonucleotide probes are formatted II
Used in the ISBH, ligate multiple probes (two or more, one probe is a fixed probe) to each other.

[0016] Still further, the present invention provides a method for detecting the presence of a target nucleic acid having a known sequence when the target is present in a sample in a very small amount compared to a homologous nucleic acid. In a preferred embodiment, the target nucleic acid is an allele that is very infrequently present in a sample having nucleic acids from multiple sources. In an alternative preferred embodiment, the target nucleic acid has a mutated sequence and is very rarely present in a sample of nucleic acids.

[0017] Still further, the invention provides a method for confirming the sequence of a target nucleic acid using single-pass gel sequencing. Primers for single-pass gel sequencing are derived from sequences obtained by SBH and these primers are used in a standard Sanger sequencing reaction to provide gel sequence information of the target nucleic acid. The sequence obtained by single pass gel sequencing is then compared to the sequence derived with SBH to confirm the sequence.

[0018] Still further, the invention provides a method for elucidating branch points using single-pass gel sequencing. Primers for single-pass gel sequencing reactions were identified from the ends of Sfs obtained after the first round of SBH sequencing, and these primers were used in a standard Sanger sequencing reaction to determine the branch points of Sfs. Gives gel sequencing information to pass. After that, the Sanger sequencing result passing through the branch point was
Align with and identify adjacent Sfs.

Further, the present invention provides a method for preparing a sample containing a target nucleic acid by PCR without purifying the PCR product before the SBH reaction. SBH format I, coarse
The PCR product can be added to the substrate without pre-purification and the substrate can be washed before introducing the labeled probe.

Further, the present invention provides a method and an apparatus for analyzing a target nucleic acid. The device comprises two arrays of nucleic acids mixed together for a desired time. In a preferred embodiment, the nucleic acids of one array are labeled. In a further preferred embodiment, a material is placed between the two arrays, which prevents mixing of the nucleic acids in the arrays. Once this material is removed or allowed to penetrate, the nucleic acids of the two arrays are mixed together. In an alternative preferred embodiment, the nucleic acids of one array are target nucleic acids and the nucleic acids of another array are oligonucleotide probes. In another preferred embodiment, the nucleic acids on both arrays are oligonucleotide probes.
In another preferred embodiment, the nucleic acids of one array are oligonucleotide probes and target nucleic acids, and the nucleic acids of another array are oligonucleotide probes. In another preferred embodiment, the nucleic acids on both arrays are oligonucleotide probes and target nucleic acids.

One method of the present invention using the apparatus described above comprises providing an array of nucleic acids immobilized on a substrate, providing a second array of nucleic acids, and an array having the nucleic acids in the second array immobilized thereon. Providing conditions to allow contact with the nucleic acid, wherein one of the arrays of nucleic acids is the target nucleic acid and the other is an oligonucleotide probe, and analyzing the hybridization results. . In a preferred embodiment, the immobilized array is a target nucleic acid and the second array is a labeled oligonucleotide probe. In a further preferred embodiment, there is some material disposed between the two arrays to prevent mixing of the nucleic acids until the material is removed or allowed to penetrate the nucleic acids.

A second method of the present invention using the above-described device provides for two arrays of nucleic acid probes and conditions that allow the two arrays of probes to contact each other and the target nucleic acid. And flanking adjacent probes on the target nucleic acid together and analyzing the results. In a preferred embodiment, 1
One array of probes is immobilized and the other array of probes is labeled.
In a further preferred embodiment, there is some material located between the two arrays,
Prevent mixing of the probe until the material is removed or allowed to penetrate the probe.

[0023] Still further, the invention provides a substrate having oligonucleotide probes immobilized thereon, wherein each probe is separated from adjacent probes by a physical barrier that resists the flow of the sample solution. In a preferred embodiment, the physical barrier is made of a hydrophobic material.

[0024] Still further, the invention provides a method of making an array of oligonucleotide probes separated by a physical barrier. In a preferred embodiment, a grid is added to the substrate using a material adding inkjet head to reduce the reaction volume of the array.

Further, the present invention provides a substrate on which an oligonucleotide is immobilized to form a three-dimensional array. The three-dimensional array has a high resolution (
Each level has a relatively low density of probes / cm ² ) and a high content of information in three dimensional space (multiple levels or probes).

[0026] Still further, the invention provides a substrate to which the oligonucleotide probe is immobilized, wherein the oligonucleotide probe has a spacer, and wherein the spacer comprises a substrate and an information portion of the oligonucleotide probe (eg, (The part of the oligonucleotide probe that binds to the target and gives sequence information). In a preferred embodiment, the spacer comprises a ribose sugar and a phosphate, wherein the phosphate covalently attaches the ribose sugar to the polymer by forming an ester with the ribose sugar via the 5 'and 3' hydroxyl groups. ing.

[0027] Still further, the invention provides a method of clustering cDNA clones into groups of similar or identical sequences so that a single representative clone can be selected from each group of sequencing. In a preferred embodiment, the method of clustering is used for sequencing a plurality of clones, wherein each clone is queried for a plurality of oligonucleotide probes;
Clustering the clones into multiple groups by determining the probes that bind to each clone and the signal strength of each probe; identifying clones that bind with similar strength to similar probes; and at least one clone from each group . In a more preferred embodiment, the plurality of probes comprises from about 50 to about 500 different probes. In another more preferred embodiment, the plurality of probes comprises about 300 different probes. In a most preferred embodiment, the plurality of clones are a plurality of cDNA clones.

[0028] Still further, the invention relates to oligonucleotide probes complexed (covalently or non-covalently) with discrete particles, wherein the particles are classified into sets based on physical properties. You. In a preferred embodiment, different probes are bound to each set of discrete particles, and the identity of the probes is determined by identifying the physical properties of the discrete particles. In an alternative embodiment, the probe is identified based on the physical properties of the probe. Physical properties include any physical property that can be used to identify discrete particles, including, for example, size, fluorescence, radioactivity, electromagnetic charge, or absorbance, or labeling a dye, radionuclide, or It can be attached to particles like EML. In a preferred embodiment, the discrete particles are separated by a flow cytometer that detects the size, charge, fluorescence, or absorbance of the particles.

The present invention also relates to a method of analyzing a target nucleic acid using a probe complexed with a discrete particle. These probes can be used in any of the methods described above with modifications to identify the probe by the physical properties of the discrete particles. These probes can also use a Format III approach, where "free" probes are identified by label and probes conjugated to discrete particles are identified by physical properties. In a preferred embodiment, the probe is SB
Used to sequence the target nucleic acid using H.

The present invention also relates to an agent that destabilizes (reduces binding energy) binding of complementary polynucleotide strands or stabilizes binding (increases binding energy) between complementary polynucleotide strands. Is related to how to use. In a preferred embodiment, the agent is an organic solvent such as a trialkylammonium salt, sodium chloride, phosphate, borate, formamide, glycol, dimethylsulfoxide, and dimethylformamide, or an amino acid analog such as urea, guanidinium, betaine. Body, polyamines such as spermidine and spermine, or other positively charged molecules that neutralize the negative charge of the phosphate backbone, detergents such as sodium dodecyl sulfate and sodium lauryl sarcosinate, minor / major groove (minor / major
groove) binders, positively charged polypeptides, and intercalating agents such as acridine, ethidium bromide, and anthracin. In a preferred embodiment, an agent is used to reduce or increase the _Tm of a pair of complementary polynucleotides. In a further preferred embodiment, the mixture of agents is used to reduce or increase the _Tm of a pair of complementary polynucleotides. In a most preferred embodiment, an agent or mixture of agents is used to improve the discrimination between complementary and perfect mismatches of complementary polynucleotides. In a preferred embodiment, one or more agents are added such that the binding energy from the AT base pair is approximately equal to the binding energy of the GC base pair. The energy of binding of these complementary polynucleotides can be increased by adding an agent that neutralizes or masks the negative charge of the phosphate groups on the polynucleotide backbone.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS SBH format 1 is suitable for the simultaneous analysis of large sample sets. Using small pieces of membrane, parallel scoring of thousands of samples on a large array can be performed in thousands of independent hybridization reactions. DNA identification can involve 1-20 probes per reaction, and in some cases, mutation identification can involve more than 1000 probes specifically selected or designed for each sample. . To identify the nature of the mutated DNA segment, it may be possible to synthesize or select a particular probe for each mutation detected in the first round of hybridization.

[0032] DNA samples can be prepared in small arrays separated by suitable spacers, and the DNA samples can be simultaneously analyzed with a probe selected from a set of oligonucleotides that can be arrayed in a multiwell plate. Can be tested. A small array can consist of one or more samples. The DNA samples in each small array can include variants or individual sequence samples. Subsequent small arrays can be organized into larger arrays. like this,
Larger arrays can include duplicates of the same small array, or can include arrays of different DNA fragment samples. The universal set of probes is D
The redundancy of reading NA fragments with predefined accuracy (eg, each base pair ("bp")
redundancy)) contains enough probes to analyze. These sets can include more probes than are needed for one particular fragment, but can contain fewer probes than are needed to test DNA samples of thousands of different sequences.

The DNA or allele identification and diagnostic sequencing process includes: 1) selecting a subset of probes from each elaborate, representative or universal set that hybridizes to each of a plurality of small arrays; 2) Adding a first probe to each sub-array of each of the arrays to be analyzed in parallel; 3) performing hybridization and scoring the hybridization results; 4) desorbing previously used probes. 5) repeating the steps of hybridization, scoring, and desorption for the remaining probes to be scored; 6) processing the obtained results to obtain additional probes to obtain a final analysis or to hybridize. 7) Perform additional hybridization for a particular subarray ; And 7) processes the complete set of data may include the steps of obtaining a final analysis.

The present approach provides for the rapid identification and sequencing of small numbers of nucleic acid samples of one type (eg, DNA, RNA), and the use of pre-synthesized, manageable-size probe sets to subarray Provides parallel analysis of many sample types of shapes. The two approaches can be combined to create an efficient and versatile process for determining DNA identity, for DNA diagnostics, and for mutant identification.

For the identification of known sequences, a smaller set of shorter probes can be used instead of longer unique probes. With this approach, more probes to score, but a universal set of probes can be synthesized to cover any type of sequence. For example, a full set of 6-mers contains only 4096 probes, and a complete set of 7-mers contains only 16,384 probes.

[0036] Complete sequencing of DNA fragments can be performed at two levels of hybridization. One level is sufficient hybridization of the probe set to cover all bases at least once. To this end, a specific set of probes for a standard sample can be synthesized. The results of hybridization with such a probe set will reveal whether and where a mutation (difference) has occurred in the non-standard sample. In addition, the probe set includes "negative" probes to confirm the hybridization results of "positive" probes. Additional specific probes can be hybridized to the sample to determine the identity of the change. This additional probe set has both "positive" (mutated sequences) and "negative" probes, and sequence changes will be identified by positive probes and confirmed by negative probes.

In another embodiment, all probes from the universal set can be scored. The universal probe set allows the scoring of a relatively small number of probes per sample in a two-step process without undesired consumption of time. The hybridization process first computes an optimal subset of hybridizing probes, and then, based on the results obtained, determines additional probes to score from a universal set of probes. It can be by two-step sequential probing. Both probe sets have "negative" probes to identify positive probes in the set. Furthermore, the sequence obtained can then be confirmed in another step by hybridizing the sample with a “negative” probe set identified from the SBH results.

In the construction of SBH sequences, special consideration should be given to K-1 oligonucleotides that are repeatedly present in the analyzed DNA fragments for opportunity or biological reasons. Without further information, relatively small fragments of DNA can be completely constructed so that every base pair is read several times.

In the construction of longer fragments, ambiguity may arise due to the repeated presence of a probe set of K-1 sequences that score positively (ie, sequences shorter than the probe length). This problem does not exist if mutations or similar sequences are determined (ie, the K-1 sequence is not identically repeated). Using knowledge of one sequence as a template, arraying positive probes and showing the best fit of the unknown sequence to the template allows sequences that are known to be similar (eg, from their presence in the database) to be similar. Can be built correctly.

The use of an array of samples can avoid continuous scoring of many oligonucleotides for one single sample or for a small set of samples. This approach allows the scoring of more probes in parallel by manipulating only one physical object.
A 1000 bp long DNA sample subarray can be sequenced in a relatively short time. If the sample is spotted on one array with 50 subarrays, the array is
Re-probing twice can score 500 probes. In screening for the presence of a mutation, sufficient probe can be used to cover each base three times. If the mutation is present, some covering probes will be affected. Using the information on the identity of the negative probe, the mutation can be mapped with a precision of two bases. An additional 15 probes can be used to elucidate the mutation of one base mapped by this method. These probes are
It covers any base combination for the two questionable positions (assuming the deletion and insertion are not involved). These probes are scored for 50 subarrays containing a given sample per cycle. In implementing a multi-label color scheme (ie, multiplexing), each has a different label, such as a different fluorescent dye.
Two to six probes can be used as a pool, thereby reducing the number of hybridization cycles and shortening the sequencing process.

In more complex cases, there may be two close mutations or insertions. They can be handled with more probes. For example, a three base insertion can be resolved with 64 probes. In the most complex case, a few steps of hybridization can be approached and a new probe set selected based on the results of previous hybridizations.

If the subarray to be analyzed contains dozens or hundreds of one type of sample, some of them may contain one or more changes (mutations, insertions or deletions). For each segment where the mutation is present, a particular probe set can be scored. The total number of probes to score for one type of sample can be hundreds. Parallel scoring of replica arrays allows scoring of hundreds of probes with relatively few cycles. In addition, compatible probes can be pooled. Positive hybridization can be assigned to selected probes to check for a particular DNA segment. This is because these segments usually differ by 75% of the constituent bases.

By using larger, longer probe sets, longer targets can be analyzed. These targets can represent a pool of fragments, such as a pool of exon clones.

The specific hybridization scoring method can be used to define the presence of a variant in the genomic segment to be sequenced from a set of diploid chromosomes. There are two variants; i) sequences from one chromosome represent known alleles;
Sequences from other chromosomes represent new mutations; or ii) both chromosomes are new but different variants. In both cases, a scanning step designed to map the change gives a two-fold maximum signal difference at the mutation location. In addition, the method can be used to identify which allele of a gene is carried by an individual and whether the individual is homozygous or heterozygous for that gene.

The scoring of the two-fold signal difference required in the first case can be performed efficiently by comparing the corresponding signal to homozygous and heterozygous controls. This approach allows one to determine the relative decrease in hybridization signal for each particular probe in a given sample. This is important because for a particular probe hybridized to a different nucleic acid fragment having the same perfect matching target, the hybridization efficiency can vary by more than two-fold. In addition, different mutation sites can affect one or more probes, depending on the number of oligonucleotide probes. Decreased signal for 2-4 consecutive probes gives more pronounced indications of the mutation site. Result is,
One or a few that can be tested and checked on a small set of selected probes, giving a perfect match signal on average 8 times stronger than the signal coming from the duplex containing the mismatch Probes are selected.

The distributed membrane allows for a very flexible experimental organization that can accommodate a relatively large number of samples representing a given sequence type or many different types of samples represented by a relatively small number of samples. become. A range of 4 to 256 samples can be handled particularly efficiently. Subarrays of dot numbers within this range can be designed to match the structure and size of standard multiwell plates used for oligonucleotide storage and labeling. The size of the subarray can be adjusted to a different number of samples, or a small number of standard subarray sizes can be used. If all samples of a type do not fit into one subarray, they can be processed with the same probe using additional subarrays or membranes. In addition, the number of replicates for each subarray can be adjusted to change the time for identification or sequencing completion.

As used herein, an “intermediate fragment” is between 5 and 10 in length.
It means an oligonucleotide of 00 bases, and preferably 10 to 40 bp in length.

In format 3, a first set of oligonucleotide probes of known sequence is immobilized on a solid support under conditions that allow them to hybridize to nucleic acids each having a complementary sequence. A labeled, second oligonucleotide probe set is provided in solution. The probes may be the same length or different lengths, both within and between sets. The nucleic acids to be sequenced or their intermediate fragments can be in double-stranded form (especially when the recA protein is present to allow hybridization under non-denaturing conditions) or single-stranded, and with varying degrees of complementarity. Can be added to the first probe set under conditions that allow for hybridisation (e.g., under conditions that allow discrimination between full and single base pair mismatched hybrids). The nucleic acids to be sequenced or their intermediate fragments can be added to the first probe set before, after or simultaneously with the second probe set. Probes that bind to adjacent sites on the target are bound together (eg, by stacking interactions or by ligase or other methods of forming a chemical bond between adjacent probes). After binding of the adjacent probes, fragments and probes that are not immobilized on the surface by chemical binding to the members of the first probe set are washed away using, for example, a high temperature (100 ° C. or lower) washing solution that melts the hybrid. The bound probe from the second set is then detected using a method appropriate for the label used (which can be, for example, a label for chemiluminescence, fluorescence, radioactivity, enzymes, densitometry, or electrophoretic mass). can do.

As used herein, nucleotide bases are “matched” or “complementary (c) if they form a stable duplex through hydrogen bonding under defined conditions.
omplementary) ". For example, under conditions commonly used in hybridization assays, adenine ("A") pairs with thymine ("T"), but not with guanine ("G") or cytosine ("C"). Do not match. Similarly, G pairs with C, but not with A or T. Inosine or Universal Base
Other bases that hydrogen bond with lower specificity, such as (“M” bases, Nichols et al., 1994), or other modified bases, such as methylated bases, may be, for example, Under conditions they are complementary to bases that form stable duplexes. A probe is formed if each base in the probe forms a duplex by hydrogen bonding with a base in the nucleic acid to form a Watson and Crick base pairing rule (ie, without any peripheral sequence effects). If the two strands are arranged by the two strands having the maximum binding energy for a particular probe, they are "perfectly complementary (
said to be "perfectly complementary" or "perfectly complementary"
match) ". "Completely complementary" and "perfect pairing"
It is meant to include probes with analogs or modified nucleotides. A "perfect match" for an analog or modified nucleotide is a "perfect matching rule" selected for the analog or modified nucleotide (e.g., having the highest binding energy for a particular analog or modified nucleotide) Binding pair). Each base in the probe that does not form a binding pair according to the "rule" is said to be "mismatched" under defined hybridization conditions.

A list of probes can be constructed where each probe is a perfect match for the nucleic acid to be sequenced. Then, the probes on this list are analyzed and arranged in the order of maximum overlap. This alignment is performed by comparing the first probe with each of the other probes on the list, where each probe has a 3 'having the longest sequence of bases identical to the sequence of the base at the 5' end of the second probe. It can be determined whether it has a terminus. Thereafter, the first and second probes can be overlapped and the process can be repeated, comparing the 5 'end of the second probe with the 3' end of all remaining probes and of the first probe. The 3 'end is compared to the 5' end of all remaining probes. The process can continue until there are no more probes on the list that do not overlap with other probes. Alternatively, one or more probes can be selected from a list of positive probes, and one or more sets of overlapping probes ("sequence nucleus") can be made in parallel. The list of probes for any such process of sequence construction can be a complete list of probes that are completely complementary to the nucleic acid to be sequenced, or can be a subset of any of them. .

The 5 ′ and 3 ′ ends of the probe overlap to extend the longer sequence extension (
stretches). This process of probe construction continues until ambiguity arises due to branch points (some probes are repeated in the fragment), repetitive sequences longer than probes, or uncloned segments. The stretch of sequence between the two ambiguities is called a fragment of the subclone sequence (Sfs). When ambiguity in sequence construction occurs due to the availability of suitable alternative overlaps with the probe, hybridization with longer probes across the site of the overlap alternative, competitive hybridization, and the terminus of the alternative Ligation of the probe with end pairs across sites, or single-pass gel analysis (to create a fuzzy alignment of Sfs) can be used.

Using the above method, from the pattern of hybridization (which can be related to the identity of the nucleic acid sample and serves as a signature to identify the nucleic acid sample) to overlapping or non-overlapping probes, ,It was constructed
Desired levels of sequence can be obtained via Sfs and up to the complete sequence for the intermediate fragment or the entire source DNA molecule (eg, chromosome).

Sequencing generally involves the following steps: (a) under conditions effective to allow the fragment to form a first complex with an immobilized probe having a complementary sequence, Array of nucleotide probes (arra
(b) contacting the nucleic acid fragment with the nucleic acid fragment; (b) the first complex hybridizes to the labeled probe, thereby forming a second complex in which the fragment is hybridized by both the immobilized probe and the labeled probe. Contacting the first complex with a set of labeled oligonucleotide probes in a solution under conditions effective to allow the binding of the first complex to the second complex; (D) detecting the presence of adjacent labeled and unlabeled probes by detecting the presence of the label; and (e) linking to the known sequence of the immobilized labeled probe. By doing so, the nucleotide sequence of the fragment is determined.

Hybridization and washing conditions are selected to detect substantially perfect hybrids (eg, those in which fragments and probes hybridize at six of seven sites, etc.). Or may be selected to allow for differences between perfect pairing and single base pair mismatches, or to allow detection of only perfect pairing hybrids.

[0055] Suitable hybridization conditions can be determined routinely by optimization procedures or experimental studies. Such procedures and reviews are routinely performed by those skilled in the art to establish protocols used in the laboratory. For example, Ausubel et al., Current Protocols i, incorporated herein by reference.
n Molecular Biology, Vol. 1-2, John Wiley & Sons (1989); Sambrook et al., Mol.
ecular Cloning A Laboratory Manual, 2nd edition, Vols. 1-3, Cold Springs Harbo
r Press (1989); and Maniatis et al., Molecular Cloning: A Laboratory Manual,
See Cold Spring Harbor Laboratory Corl Spring Harbor, New York (1982). For example, conditions such as temperature, component concentrations, hybridization and washing times, buffer components, and their pH and ionic strength can be varied.

In embodiments where the labeled immobilized probe is not physically or chemically attached, detection may rely solely on controlled stringency washing steps. Under such conditions,
Adjacent probes are used for stacking interactions between adjacent probes.
ons) and has a high binding affinity. Conditions can be varied to optimize the process as described above.

In embodiments where the immobilized labeled probe is ligated, the linkage may be a chemical ligation reagent (
Water-soluble carbodiimide or cyanogen bromide),
Or a ligase enzyme, can be utilized commercially available T ₄ DNA ligase. Washing conditions utilize the difference in stability of non-adjacent probes relative to adjacent probes,
A choice can be made to distinguish non-adjacent labeled immobilized probes relative to adjacent labeled immobilized probes.

The oligonucleotide probes can be labeled with a fluorescent dye, a chemiluminescent system, a radioactive label (eg, ³⁵ S, ³ H, ³² P or ³³ P) or an isotope that can be detected by mass spectrometry.

If a nucleic acid molecule of unknown sequence is longer than about 45 or 50 bp, the molecule can be fragmented and the fragment sequenced. Fragmentation can be accomplished by restriction digestion, shearing or
Can be achieved by NaOH. The fragments can be separated in size (eg, by gel electrophoresis) to obtain a preferred fragment of about 10-40 bp in length.

[0060] Oligonucleotides can be prepared in a number of ways known to those skilled in the art, for example, nucleoside phosphoramidites or hydrogenated nucleosides.
Laser-activated photodeprotection attachment via a phosphate group using reagents such as ide hydrogen phosphorate
)). Glass, nylon, silicon and fluorinated carbon supports can be used.

Oligonucleotides are incorporated into arrays, which can include all or a subset of all probes of a given length, or a set of probes of a selected length. Hydrophobic partitions can be used to separate probes or subarrays of probes. Arrays can be designed for a variety of applications (eg, mapping, partial sequencing, sequencing of target regions for diagnostic purposes, mRNA sequencing, and bulk sequencing). Specific chips can be designed for specific applications by choosing the combination and sequence of probes on the substrate.

For example, 1024 immobilized probe arrays of all oligonucleotide probes of 5 base pairs in length (each array containing 1024 different probes) can be constructed. The probes in this example are 5-mers in the informative sense (they may actually be longer probes). A second set of 1024 5-mer probes is labeled, and one of each labeled probe can be applied to an array of immobilized probes with the fragment to be sequenced. In this example, 1024 arrays would be combined into a large superarray, or "superchip." In those instances where the labeled probe and one of the labeled probes hybridize end-to-end along the nucleic acid fragment, the two probes are combined, for example, by a ligation reaction to remove unbound label. Later, a 10-mer complementary to the sample fragment is detected by correlation of the presence of the label at a point in the array with the known sequence of immobilized probes to which the known sequence of the labeled probe is applied. The sequence of the sample fragment is simply the sequence of the immobilized probe contiguous with the sequence of the labeled probe. Thus, all one million possible 10-
The mer uses only the 5-mer, and thus has a thousandth of the labor of oligonucleotide synthesis.
It can be tested by a combinatorial process involving 1.

In a preferred embodiment, the substrate supporting the array of oligonucleotide probes is partitioned such that each probe in the array is separated from adjacent probes by a physical barrier, which can be, for example, a hydrophobic material. Can be In a preferred embodiment, the physical barrier has a width between 100 μm and 30 μm. In a more preferred embodiment, the distance from the center of each probe to the center of any adjacent probe is 325 μm. These probe arrays have a nonmoving immobilized substrate or inkjet deposition device, such as a microdroplet head, and a suitable robotic system, such as an anorad gantry. It can be "mass-produced" using a substrate fixed to a rotating drum or plate.

[0064] In another preferred embodiment, the oligonucleotide probes are immobilized on a three-dimensional array. A three-dimensional array is composed of a number of layers, each of which can be analyzed separately from the other layers. A three-dimensional array can take a number of forms, including, for example, the following forms: a probe can be located on a substrate having a number of depressions located at different depths within the depressions (each plane is The array is composed of probes of similar depth within the cavity); or the array has different depths where the probe is located at the bottom of the cavity, or at a peak that divides the cavity or a combination of several peaks. Either placed on a substrate with depressions, depressions can be used (each plane is composed of all probes at a constant depth); or the array is stacked 3
It can be placed on a substrate composed of multiple sheets forming a dimensional array.

The probes in these arrays can include spacers that increase the distance between the substrate surface and the information providing portion of the probe. Spacers are carbon, silicon, oxygen,
Sulfur, may be composed of atoms capable of forming at least a divalent bond such as phosphorus, or a sugar-phosphate group, amino acid, peptide, nucleoside, nucleotide, sugar, hydrocarbon, aromatic ring, hydrocarbon ring, linear And a molecule capable of forming at least a divalent bond such as a branched hydrocarbon.

The nucleic acid sample to be sequenced can be fragmented or otherwise processed (eg, recA).
To avoid interference with hybridization by the second structure in the sample. The sample can be fragmented, for example, by digestion with a restriction enzyme such as Cvi, physical cleavage (eg, by sonication), or NaOH treatment. The resulting fragments can be separated by gel electrophoresis, and fragments of appropriate length, such as between about 10 bp and about 40 bp, can be extracted from the gel. In a preferred embodiment, "fragments" of the nucleic acid sample cannot be ligated to other fragments in the pool. A pool of such fragments is obtained by treating the fragmented nucleic acid with a phosphatase (eg, calf intestinal phosphatase). Alternatively, non-ligated fragments of a sample nucleic acid can be prepared by using random primers (eg, N ₅ -N ₉ , where N = A, G, T, or C) in a Sanger-deoxy sequencing reaction with the sample nucleic acid. Is obtained. This will produce a DNA fragment that has a sequence complementary to the target nucleic acid and is terminated by a dideoxy residue that cannot be ligated to another fragment.

Reusable Format 3 SBH arrays can be made by introducing a cleavable bond between the immobilized probe and the labeled probe, and then cleaving this bond after the round of Format 3 analysis has been completed. Labeled probes are ribonucleotides or ribonucleotides and can be used as binding bases in labeled probes, which can be subsequently removed by, for example, RNAse or uracil-DNA glycosylation, or NaOH treatment. In addition, bonds made by chemical ligation can be selectively cleaved.

Other variations include the use of modified oligonucleotides to increase specificity or efficiency, such as conditions optimally selected for the first set of labeled probes (eg,
Temperature), followed by a second hybridization of the labeled probe.
Includes cycling hybridization that increases the hybridization signal by hybridizing under conditions optimally selected for the set. The reading frame shift can be determined by using a mixture (preferably an equimolar mixture) of probes ending with each of the four nucleotide bases A, T, C and G.

Branch points create ambiguity regarding the ordered arrangement of fragments. Sequence information is SB
If there is such an ambiguity (“branch point”), as determined by H, then to organize the hybridization data, (i) the cost fraction of complete gel sequencing Long read length single pass gel sequencing; or (ii) comparison with related sequences can be used. Primers for single pass gel sequencing through branch points were identified from SBH sequence information or from known vector sequences, e.g., sequences adjacent to the vector insertion site, and used for standard Sanger-sequencing reactions on sample nucleic acids. Do. The sequence obtained from this single pass gel sequencing was read into and out of the branch point.
the branch points), and compare the Sfs order with Sfs to identify. Alternatively, Sfs can be aligned by comparing the sequence of Sfs to related sequences, aligning Sfs, and producing the sequence closest to the related sequence.

Furthermore, the number of tandemly repeated nucleic acid segments in a target fragment can be determined by single-pass gel sequencing. Because tandem repeats occur infrequently in the protein-coding portion of a gene, the gel sequencing step is performed when one of these non-coding regions is identified as of particular interest (eg, it is an important regulatory region). Case) will only take place.

By obtaining information about the degree of hybridization exhibited by a set of only about 200 oligonucleotide probes, the unique signature of each gene (si
gnature) can be used to classify cDNA from the library and determine whether the library contains multiple copies of the same gene. Such signatures can distinguish identical, similar and different cDNAs and create an inventory.

[0072] Methods for isolating, cloning and sequencing nucleic acids and nucleic acids are well known to those skilled in the art. For example, Ausubel et al., Current, which is incorporated herein by reference.
Protocols in Molecular Biology, Vol. 1-2, John Wiley & Sons (1989); and Sambrook et al., Molecular Cloning A Laboratory Manual, 2nd edition, Vols. 1-3, C
See old Springs Harbor Press (1989).

[0073] SBH is a highly developed technology that can be implemented by a number of methods known to those skilled in the art. In particular, techniques related to sequencing by hybridization in the following references are incorporated herein by reference: Drmanac et al., April 1993.
U.S. Patent No. 5,202,231, issued 13th, which is hereby incorporated by reference; Drmanac et al., Genomics, 4, 114-128 (1989), Drmanac et al., Proceeding.
s of the First Int'l.Conf.Electrophoresis Supercomputing Human Genome,
Edited by Cantor et al., World Scientific Pub. Co., Singapore, 47-59 (1991); Drman
ac et al., Science, 260, 1649-1652 (1993); Lehrach et al., Genome Analysis: Geneti.
c and Physical Mapping, 1, 39-81 (1990), Cold Spring Harbor Laboratory P
ress; Drmanac et al., Nucl. Acids Res., 4691 (1986); Stevanovic et al., Gene, 79 1
39 (1989); Panusku et al., Mol. Biol. Evol., 1, 607 (1990); Nizetic et al., Nucl.
Acids Res., 19 182 (1991); Drmanac et al., J. Biomol. Struct. Dyn., 5, 1085 (
1991); Hoheisel et al., Mol. Gen., 4, 125-132 (1991); Strezoska et al., Proc. Nat '.
l. Acad. Sci. (USA), 88, 10089 (1991); Drmanac et al., Nucl. Acids Res., 19,
5839 (1991); and Drmanac et al., Int. J. Genome Res., 1, 59-79 (1992).

The present invention is described in the following examples. In view of the disclosure herein, one skilled in the art will recognize that many other embodiments and variations can be made within the scope of the present invention. Therefore, the broader aspects of the present invention are not intended to be limited to the disclosure of the following examples.

Example 1 Preparation of a Probe Set A universal set of two types of probes can be prepared. The first is a complete set (or at least a non-complementary subset) of relatively short probes, such as a total of 4096 (or about 2000 non-complementary) 6-mers, or a total of 16,384 (or about 8,000 non-complementary). Complementary)
7-mer. All non-complementary 8-mer subsets and longer probes have 32
Less convenient because it contains more than 2,000 probes.

[0076] The second type of probe set is selected as a small subset of probes that is still sufficient to read any bp (base pairs) of any sequence by at least one probe. For example, 12 out of 16 2-mers are sufficient. A small subset of 7-mers, 8-mers and 9-mers for sequencing double-stranded DNA can be about 3,000, 10,000 and 30,000 probes, respectively.

[0077] Probe sets can also be selected to identify target nucleic acids of known sequence and / or to identify alleles or mutants of the target nucleic acid with known sequences. Such a probe set contains sufficient probes so that any nucleotide position of the target nucleic acid is read at least once. Alleles or mutants are identified by the loss of binding of one of the "positive" probes. The specific sequences of these alleles or mutants are then interrogated with the target nucleic acid and probe sets, including any possible nucleotide exchanges and combinations of exchanges at these probe positions.
g).

The probe set can also consist of from 50 probes to a universal set of probes (all probes of a specific length), more preferably the set is 100-50
Consisting of zero probes, in a most preferred embodiment, the probe set is 300
Probes. In a preferred embodiment, the probe set is 6-9 nucleotides in length, and the cDNA clones are assembled into groups of similar or identical sequences (cl
a single representative clone can be selected from each group for sequencing.

Probes can be prepared using standard chemistry with 1-3 non-specific (mixed A, T, C and G) or terminal universal (eg, M base or inosine) bases. . If a radiolabel is used, the probe can have an OH group at the 5 'end for substrate kinasing by the radiolabeled phosphate group. Alternatively, any compatible system can be used, such as a probe labeled with a fluorescent dye or the like. Other types of probes may also be used, such as PNAs (protein nucleic acids), or probes containing modified bases that alter duplex stability.

The probes can be stored in multiple well plates with bar codes. For a small number of probes, a 96-well plate can be used; for 10,000 or more probes, 38
Storage in 4- or 864-well plates is preferred. With a stack of 5-50 plates,
Enough to store all probes. Approximately 5 pg of probe may be sufficient for hybridization with one DNA sample. Thus, from a small synthesis of about 50 mg per probe, 10 million samples can be analyzed. If each probe is always used for a third sample and each sample is 1000 bp in length, then more than 30 billion bases (10 human genomes) are sequenced with 5,000 probe sets. Can be done.

Example 2 Probes Having Modified Oligonucleotides A modified oligonucleotide can be introduced into a hybridization probe and used under appropriate conditions. For example, C ⁵ - pyrimidines having a halogen in position, the duplex stability by affecting the base stack (stack) can be used to improve. 2,6-Diaminopurine can be used to provide a tertiary hydrogen bond for base pairing with thymine, thereby thermally stabilizing the DNA duplex. The use of 2,5-diaminopurine increases duplex stability and allows for more stringent conditions of annealing, thereby improving the specificity of duplex formation and reducing background problems, Allows for the use of shorter oligomers.

The synthesis of tri-phosphorylated forms of these modified nucleotides is described in Hoheisel & Lehrach (1
990).

[0083] Non-discriminatory base analogs, or universal bases, designed by Nichols et al. (1994) can also be used. This new analog, 1- (2-deoxy-D-ribofuranosyl) -3-nitropyrrole (referred to as M), results from the degeneracy of the gene code or only fragmented peptide sequence data. Created for use in oligonucleotide probes and primers to eliminate design issues when not available. This analog maximizes stacking while minimizing hydrogen binding interactions without sterically disrupting the DNA duplex.

The M nucleoside analogs are designed to minimize stacking interactions using neutral polar substituents attached to the heteroaromatic ring to enhance intrachain and interchain stacking interactions to enhance base pairing. Reduced the role of hydrogen bonding. (1994) encouraged 3-nitropyrrole 2-deoxyribonucleoside because of its structural and electronic similarity to p-nitroaniline, a derivative of which is one of the smallest known inserts of double-stranded DNA. did.

The dimethoxytrityl protected phosphoramidites of nucleoside M can also be used to incorporate into nucleotides used as primers for sequencing and polymerase chain reaction (PCR). Nichols et al. (1994) have shown that a significant number of nucleotides can be replaced by M without reducing the specificity of the primer.

The unique property of M is that it replaces long strings of contiguous nucleosides,
Also, its ability to generate functional sequencing primers. 3, 6 and 9 M
Sequences with substitutions are all readable sequencing ladders (ladd
ers), and PCR with three different M-containing primers
All resulted in the amplification of the correct product (Nichols et al., 1994).

The ability of the 3-nitropyrrole-containing oligonucleotide to function as a primer strongly suggests that the double-stranded structure must be formed by complementary strands. Oligonucleotide pair _{d (5-C 2 -T 5} XT 5 G 2 -3) and _{d (5-C 2 A 5} YA 5 G 2 -3
) (Where X and Y can be A, C, G, T or M), the optical thermal profile obtained for DNA double-stranded to single-stranded transitions Was reported to be consistent with the normal sigmoid pattern observed. XM base pairs (where X is
A, C, G, or T, and Y is M).
It was reported to fall within the 3 ° C. range (Nichols et al., 1994).

Example 3 Probe Selection and Labeling Once an array of subarrays has been created, a set of probes to be hybridized to each subarray during each hybridization cycle is defined. For example, 38
Four probe sets can be selected from the universal set, and 96 probings can be performed in each of four cycles. Probes selected to hybridize in one cycle preferably have similar G + C content.

Transfer the probes selected for each cycle to a 96-well plate and then label them by kinasing if they have not been labeled (eg, with a stable fluorescent dye) before storage. Or by other labeling procedures.

Based on the first round of hybridization, a new set of probes can be defined for each subarray for further cycles. Some arrays cannot be used in some cycles. For example, if only eight out of 64 patient samples show mutations and eight probes are evaluated initially for each mutation, then all 64 probes will be in one cycle. Evaluated, 32 subarrays not used. These unused subarrays may be treated with a hybridization buffer to prevent the filters from drying out.

Probes can be prepared from the storage plate by any convenient approach, such as a single channel pipetting device, or a robotic station such as the Beckman Biomek 100
0 (Beckman Instruments, Fullerton, California) or Mega Two Robot (Meg
amation, Lawrencevill, New Jersey). The robot station can be integrated with a data analysis program and a probe management program. The output of these programs can be input to one or more robot stations.

The probes can be collected one by one and added to a subarray covered by a hybridization buffer. Preferably, the collection probe is placed on a new plate, labeled or mixed with a hybridization buffer. A preferred method of recovery is by accessing one storage plate at a time and pipetting (or transferring by metal pins) a sufficient amount of each selected probe from each plate to a particular well in the relay plate. Arrays of individually addressable pipettes or pins can be used to speed up the collection process.

Example 4 Preparation of Labeled Probes Oligonucleotide probes can be prepared by automated synthesis, and are routine for those skilled in the art, for example, using the system of Applied Biosystems. Alternatively, probes can be manufactured using the Genosys Biotechnologies Inc. method using a stack of porous Teflon wafers.

Oligonucleotide probes can be radiolabeled (eg, ³⁵ S, ³² P, preferably ³³ P) for arrays with 100-200 μm or 100-400 μm spots; non-radioactive isotopes (Jacobsen et al.). , 1990); or a fluorophore (Brumbaugh et al., 1988). All such labeling methods are described by the appropriate sections in Sambrook et al. (1989) and by Schubert et al. (1990), Murakami et al. (1991) and Cate et al. (1991).
) Are routine in the art, as exemplified by additional references such as) and all of these references are specifically incorporated herein by reference.

For radiolabeling, common methods are end labeling using T4 polynucleotide kinase or high specificity radiolabeling using Klenow or T7 polymerase.
These are described below.

Synthetic oligonucleotides are synthesized without a phosphate group at their 5 ′ end, and therefore use the enzyme bacteriophage T4 polynucleotide kinase [ ^-32 P].
ATP or [- ³³ P] from ATP - ³² P or - ³³ are easily labeled by transfer of P. Provided that the reaction is carried out efficiently, the specific activity of such probes, the [- ³² P] ATP or [- ³³ P] can be as high as ATP itself specific activity. The following reaction is 10 pm
Designed to label ole oligonucleotides with high specific activity. Labeling of different amounts of oligonucleotide can be easily achieved by increasing or decreasing the size of the reaction and keeping the concentration of all components constant.

[0097] The reaction mixture, 10 [mu] l of the oligonucleotide (10 pmole / μl); 10 × bacteriophage T4 polynucleotide kinase buffer 2.0 .mu.l; of 5.0μl [- ³² P] AT
P or [- ³³ P] ATP (specific activity 5000 Ci / mmole; 10 mCi / ml in aqueous solution) (10 pmole);
And 11.4 μl of water. Eight units (about 1 μl) of bacteriophage T4 polynucleotide kinase are added to the reaction mixture and incubated at 37 ° C. for 45 minutes. The reaction is heated at 68 ° C. for 10 minutes to inactivate bacteriophage T4 polynucleotide kinase.

The transfer efficiency of ³² P or ³³ P to the oligonucleotide and its specific activity are then determined. If the specific activity of the probe is acceptable, purify it. If the specific activity is too low, add an additional 8 units of enzyme and add 6 reactants to inactivate the enzyme.
Incubate at 37 ° C for another 30 minutes before heating at 8 ° C for 10 minutes.

Purification of the radiolabeled oligonucleotide includes, for example, precipitation with ethanol; precipitation with cetylpyridinium bromide; chromatography on Bio-Gel P-60;
Alternatively, it can be achieved by chromatography on a Sep-Pak _C18 column, or by polyacrylamide gel electrophoresis.

Probes with higher specific radioactivity are obtained by using the Klenow fragment of E. coli DNA polymerase I to synthesize a DNA strand complementary to the synthetic oligonucleotide. A short primer is used to complement the radiolabeled probe whose sequence is desired.
) Is hybridized to the oligonucleotide template. Then, the primer was used as a template-directing method (template-direc) using the Klenow fragment of E. coli DNA polymerase I.
In ^{ted manner) [- 32 P]} dNTP or [- ³³ P] extended by introducing dNTPs. After the reaction, the template and the product are denatured and separated by electrophoresis on a polyacrylamide gel under denaturing conditions. According to this method, it is possible to produce an oligonucleotide probe containing several radioactive atoms per oligonucleotide molecule.

To utilize this method, a sufficient amount of [a- ³² P] dNTP or [a- ³³ P] is used to achieve the desired specific activity and complete the synthesis of all template strands. The dNTPs are mixed in a microcentrifuge tube. Then, appropriate amounts of primer and template DNA are added to the tube in a 3 to 10-fold molar excess of the primer over the mold.

Next, 0.1 volume of 10 × Klenow buffer is added and mixed well. Then, 2
Add -4 units of the Klenow fragment of E. coli DNA polymerase I in 5 μl reaction volumes, mix and incubate at 4 ° C. for 2-3 hours. If desired, the progress of the reaction is monitored by a small amount of (
0.1 μl) aliquots can be removed and monitored by measuring the percentage of radioactivity that can be precipitated by 10% trichloroacetic acid (TCA).

The reaction is diluted with an equal volume of gel-loading buffer, heated to 80 ° C. for 3 minutes, and then the entire sample is loaded on a denaturing polyacrylamide gel. Following electrophoresis, the gel was subjected to autoradiography,
The probe is localized and removed from the gel. Various methods of fluorescent probe labeling are also available, for example, Brumbaugh et al. (1988) describe the synthesis of fluorescently labeled primers. A deoxyuridine analog is synthesized having a "linker arm" of a primary amine consisting of 12 atoms attached to the C-5 position. The synthesis of analogs consists of deriving 2-deoxyuridine via an organometallic intermediate to give 5- (methylpropenyl) -2-deoxyuridine. Reaction with the dimethoxytrityl chloride product produces the corresponding 5-dimethoxytrityl addition product. The methyl ester is hydrolyzed, activated and reacted with an appropriately monoacylated alkyl diamine. After purification, the resulting linker arm nucleoside is converted to a nucleoside analog suitable for chemical oligonucleotide synthesis.

[0104] Oligonucleotides containing one or two linker arm bases are then made using modified phosphoridite chemistry. To a solution of 50 nmol linker arm oligonucleotide in 25 μl of 500 mM sodium carbonate (pH 9.4), add 20 μl of 300 mM FITC in dimethyl sulfoxide. The mixture is stirred at room temperature for 6 hours. Oligonucleotides are separated from free FITC by elution from a 1 x 30 cm Sephadex G-25 column with 20 mM ammonium acetate (pH 6) and combining the fractions of the first UV absorption peak. .

In general, fluorescent labeling of an oligonucleotide at its 5 ′ end involves two steps. First, an N-protected aminoalkyl phosphoramidite derivative is added to the 5 'end of the oligonucleotide during automated nucleic acid synthesis. After removal of all protecting groups, the NHS ester of the appropriate fluorescent dye was coupled with the 5'-amino group overnight and then reversed phase H
Purify the labeled oligonucleotide from excess dye using PLC or PAGE.

[0106] Schubert et al. (1990) reported the synthesis of phosphoramidites, which allows oligonucleotides labeled with fluorescein to be produced during automated DNA synthesis.

Murakami et al. Also reported the production of fluorescein-labeled oligonucleotides. Cate et al. (1991) report the use of oligonucleotide probes that are directly conjugated to alkaline phosphatase in combination with a direct chemiluminescent substrate (AMPPD) that allows probe detection.

Labeled probes will be readily available for purchase from a variety of commercial sources, including GENSET, rather than being synthesized.

Other labels include ligands that can serve as members that specifically bind to the labeled antibody, chemiluminescers, enzymes, antibodies that can serve as specific binding pair members for the labeled ligand, and the like. A wide variety of labels have been used in readily available immunoassays. Other labels include antigens, groups with specific reactivity, and components that are electrochemically detectable.

In general, labeling of nucleic acids by electrophore mass labels (EML) is described, for example, in Xu et al., J. Chromatography 764: 95-102 (1997). Electrophores are compounds that can be detected with high sensitivity by electron capture mass spectrometry (EC-MS). EML is a well-known chemistry in the art for reversibly modifying nucleotides (eg, well-known nucleotide synthesis chemistry teaches a variety of methods for attaching molecules to nucleotides as protecting groups). Can be used to bind to the probe. EML is detected using a variety of well-known electron capture mass spectrometers (eg, those sold by Finnigan Corporation).
In addition, techniques that can be used to detect EML include, for example, fast atom bombardment (f
ast atomic bombardment) mass spectrometry (eg, Koster et al., Biomedical Environ,
Mass Spec. 14: 111-116 (1987)); plasma desorption mass spectrometry; electrospray / ion spray method (eg, Fenn et al., J. Phys. Chem. 88: 4451-59 (198)
4), International Application No. WO 90/14148, Smith et al., Anal. Chem. 62: 882-89 (1990)); and matrix-assisted laser desorption / ionization method (Hillenkamp et al., "M.
atorix Assisted UV-Laser Desorption / Ionization: A New Approach to Mass S
pectrometry of Large Biomolecules. "Biological Mass Spectrometry (Burlin
game and MacCloskey), Elsevier Science Publishers, Amsterdam, pp. 49
Huth-Fehre et al., "Matrix Assisted Laser Desorption Mass Spectr.
ometry of Oligodeoxythymidylic Acids ", Rapid Communications in Mass Spec
trometry, 6: 209-13 (1992)).

[0111] In a preferred embodiment, the EML is linked to the probe by a light-sensitive covalent bond. After hybridization with the target nucleic acid, the EML is released from the probe by a laser or other light source that emits light of the desired wavelength. EML then G
It is sent to a C-MS (gas chromatograph-mass spectrometer) or other suitable device and identified by its mass.

Example 5 Preparation of Sequencing Chips and Arrays In a basic example, a 50 micron chip was used to obtain a 3 × 3 mm chip that could be combined to obtain a 20 × 20 cm array. Use a 6-mer attached to the surface. In another embodiment, a 9 × mer chip of 5 × 5 mm size (4000 units of such a chip is used to make a 30 × 30 cm array) bonded to a 10 × 10 micron surface A 9-mer oligonucleotide is used. Arrays containing 4,000 to 16,000 oligo chips are arranged in a square array. The plate or tube collection, as depicted, can be packaged with the array as part of a sequencing kit.

The arrays can be separated from each other physically or by hydrophobic surfaces. One possible way to utilize hydrophobic strip separation is to use QA Laboratories, Tr.
using technologies such as the Iso-Grid Microbiology System manufactured by onto, Canada.

[0114] Hydrophobic grid membrane filters (HGMF) have been used in analytical food microbiology for about 10 years if they exhibit the unique attraction of extended numerical range and automatic colony counting. Was. One commercially available grid is a polysulfone polymer (Gelman Tuffryn HT-450, 0.45μ pore size) on which is printed a black hydrophobic ink grid of 1600 (40 × 40) square cells. ) Square (60 x 60 cm) QA Laboratories Ltd. (Toronto,
Canada) ISO-GRID®. HGMF is pre-inoculated with the bacterial suspension by suction filtration and incubated on a choice of different or selective media.

The HGMF function is more similar to an MPN device than a conventional plate or membrane filter because microbial growth is limited to grid cells with known locations and sizes on the membrane. Peterkin et al. (1987) suggest that when used with an HGMF replicator, these
We reported that HGMF can be used to expand and store genomic libraries.
One such device replicates the growth from each of the 1600 cells of ISO-GRID, allowing many copies of the master HGMF to be made (Peterkin et al., 1987).

[0116] Sharpe et al. (1989) also reported that ISO-GRID HGMF and Auto H
A GMF counter (MI-100 interpreter) and RP-100 replicator were used. They reported techniques for maintaining and screening many microbial cultures.

Peterkin and colleagues later reported a method for screening DNA probes using a hydrophobic grid membrane filter (Peterkin et al., 1989). These authors, HGM
An efficient direct colony hybridization method on F was reported. Previously, poor results were obtained due to the low DNA binding capacity of the epoxy sulfone polymer on which HGMF was printed. However, Peterkin et al. (1989)
HG replicated and incubated before binding of DNA to the membrane surface contacted the DNA.
It was reported that MF was improved by treating with polyethyleneimine and polycation. This early work uses cellular DNA binding and has a different purpose than the present invention, but the method described is readily applicable to format 3 SBH.

To quickly identify useful sequences, Peterkin et al. (1989) used radiolabeled plasmid DNA from various clones and tested its specificity for DNA on HGMF produced. In this method, DNA from recombinant plasmids was rapidly screened by colony hybridization to 100 microorganisms on HGMF, which can be easily and reproducibly produced.

Operation with small (2-3 mm) tips and parallel runs of thousands of reactions. The solution of the present invention is to keep the chips and probes in a corresponding array. 1
In one example, a chip containing 250,000 9-mers was placed in an 8 × 8 mM plate (15 μM / oligonucleotide, 15 μM / oligonucleotide, aligned in an 8 × 12 format (96 chips) with a 1 mM groove in the middle. (Pease et al., 1994). Probes are added, one probe on one chip, either by a multi-channel pipette or a pin array. To evaluate all 4000 6-mers, 42 chip arrays must be used, either using different ones or reusing one set of chip arrays many times. Must.

In the above case, using the original nomenclature of the present application, F = 9; P = 6; and F + P = 15
It is. The chip is represented by the formula BxNn, where x is the number of specific bases B; n is the number of non-specific bases, so that x = 4-10 and n = 1-4. It can have a probe. To achieve more efficient hybridization and to avoid the potential effects of any support oligonucleotide, certain bases can be surrounded by non-specific bases, and thus (N) nBx ( N) m and the like.

In another embodiment of the chip, the substrate supporting the array of oligonucleotide probes is partitioned into sections, each probe in the array being adjacent by a physical barrier, which can be, for example, a hydrophobic material. Are separated from the probe. In a preferred embodiment, the physical barriers have a width between 300 μm and 30 μm, and the distance between the center of each physical barrier and the center of the next physical barrier is at least 32
5 μm.

In a preferred embodiment, the hydrophobic material is placed on a substrate and forms a barrier of a desired width using an inkjet head, coupled to a suitable robotic system. For example, a microdroplet head adapted to apply a suspension or solution of a desired hydrophobic material (eg, an oil-based material that forms a barrier after the solvent evaporates) is an anodrad gantry. A) a grid of hydrophobic material can be applied on the desired substrate to form multiple wells on the substrate, which can be combined with a suitable housing and dispensing system. After forming a grid of hydrophobic material, a different probe was placed on each well using a robotic system similar to that used for grid formation but applied to apply the solution or suspension of the probe. (Or a mixture of probes is applied to each well). In one embodiment, the same robotic system is used to apply the hydrophobic grid and probe. In this embodiment, the dispensing system is flushed with water after the hydrophobic grid has been applied and then primed for probe delivery.

Example 6 Preparation of Support-Bound Oligonucleotides Oligonucleotides, ie, small nucleic acid segments, are synthesized directly by chemical means, eg, by chemical means, commonly performed using, for example, an automated oligonucleotide synthesizer. Easy to manufacture.

In general, oligonucleotides can be attached to a support by a suitable reactive group.
Such groups are well known in the art, for example, amino (-NH _2); containing or carboxyl (-CO ₂ H) group; hydroxylation (-OH). Oligonucleotides attached to a support can be prepared by any method known to those skilled in the art using any suitable support, such as glass, polystyrene or Teflon. One method is to accurately spot oligonucleotides synthesized by a standard synthesizer. Immobilization can be performed, for example, using passive adsorption (Inouye &
Hondo, 1989): Using UV light (Nagata et al., 1985; Dahlen et al., 1987; Morriey and Collins, 1989): or by covalent attachment of base-modified DNA (Keller et al., 1988).
1989); or by formation of an amide group between the probe and the support (Wall et al., 1995;
Chebab et al., 1992; and Zhang et al., 1991); can be achieved by a number of methods, including the method of which all references are specifically incorporated herein.

Another method that can be used is the use of strong biotin-streptavidin interaction as a linker. For example, Broude et al. (1994) report the use of biotinylated probes immobilized on streptavidin-coated magnetic beads, although they are double-stranded probes. Streptavidin coated beads are available from Dynal, Osl
This same conjugation chemistry, of course, can be purchased from o, but can be applied to coat any surface with streptavidin. Biotinylated probes can be purchased from various sources, such as, for example, Operon Technologies (Alameda, CA).

Nunc Laboratories (Naperville, IL) also sells suitable materials that can be used. Nunc Laboratories has developed a method by which DNA can be covalently attached to a microwell surface called Covalink NH. Kovarin NH is a polystyrene surface grafted with secondary amino groups (> NH) that serves as a crosslinking head for further covalent attachment. CovaLink Modules are
Available from Nunc Laboratories. DNA molecules can be exclusively attached to Kovarin at the 5 'end by phosphoramidate linkage, allowing immobilization of more than 1 pmol of DNA (Rasmussen et al., 1991).

The use of Kovarinc NH strips for covalent attachment of DNA molecules at the 5 'end has been reported (Rasmussen et al., 1991). This technique utilizes phosphoramidite linkages (Chu et al., 1983). This is advantageous because immobilization using only a single covalent bond is preferred. The phosphoramidite linkage attaches the DNA to the secondary amino group of Kovarin NH, which is located at the end of a spacer arm that is covalently grafted onto the polystyrene surface by a 2 nm long spacer arm. In order for the oligonucleotide to be linked to Kovarin NH via a phosphoramidate linkage, the oligonucleotide terminus must have a 5 'terminal phosphate group. perhaps,
It would even be possible for biotin to be covalently linked to Kovalin and then streptavidin used to bind the probe.

More specifically, the binding method involves dissolving the DNA in water (7.5 ng / μl), denaturing at 95 ° C. for 10 minutes, and cooling on ice for 10 minutes. Then, ice-cooled 0.1 M 1-methylimidazole, pH 7.0 (1-MeIm ₇ ) is added to a final concentration of 10 mM 1-Melm ₇ . The ss DNA solution is then dispensed onto Kovarink NH strips placed on ice.

A fresh preparation of carbodiimide, 0.2 M 1-ethyl-3- (3-dimethylaminopropyl) -carbodiimide (EDC), dissolved in 10 mM 1-Melm ₇ , was made up to 2 per well.
Add 5 μl. Incubate the strip at 50 ° C. for 5 hours. After the incubation, the strips are for example washed with Nunc-Immuno Wash.
Wash using). That is, first wash the wells three times and then wash with
Immersion for 3 min and finally 3 washes (where the wash solution is 0.4 N NaO heated to 50 ° C)
H, including 0.25% SDS).

Further suitable methods for use according to the invention are believed to be those described in PCT Patent Application No. WO 90/03382 (Southern and Maskos), which is incorporated herein by reference. A method for producing a support-bound oligonucleotide comprises coupling a nucleoside 3'-reagent to a support-supported aliphatic hydroxyl group via a phosphate group via a covalent phosphodiester bond. next,
Oligonucleotides are synthesized on a supported nucleoside and remove protecting groups from the synthetic oligonucleotide chain under standard conditions that do not cleave the oligonucleotide from the support. Suitable reagents include nucleoside phosphoramidites and nucleoside hydrogen phosphorates.

An on-chip method for producing DNA probes for the production of DNA probe arrays (on
-chip strategy) is available. For example, addressable laser-activated photodeprotection is described in Fod, incorporated herein by reference.
(1991) can be used in direct oligonucleotide chemical synthesis on glass surfaces. Probes may also be immobilized on nylon supports as reported by Van Ness et al. (1991); or Duncan and Cavalier (
1988) can also be conjugated to Teflon using all of these references.
It is specifically incorporated herein.

The binding of the oligonucleotide to the nylon support, reported by Van Ness et al. (1991), activated the nylon surface by alkylation and the selective activation of the 5'-amine of the oligonucleotide by cyanuric chloride. I need.

One particular method for making support-bound oligonucleotides is the light-generated method reported by Pease et al. (1994), which is incorporated herein by reference. The use of composition. These authors use current photolithographic techniques to create immobilized oligonucleotide probe arrays (DN
A chip). These methods of using light to direct the synthesis of oligonucleotide probes in high-density microarrays are based on photolabile 5'-protected N-acyl-deoxynucleoside phosphoramidites, surface linker chemistry and versatile combinatorial synthesis. Take advantage of the method. A matrix of 256 spatially defined oligonucleotide probes can be made in this way and used in the advantageous Format 3 sequencing described herein.

Of course, DNA chips, such as one of the photoactivated chips described above, would be readily available from commercial sources. In this regard, see Affymet of Santa Clara, CA 95051.
You can contact riox and Beckman.

In a preferred embodiment, the probes of the present invention comprise an information providing moiety (a moiety that hybridizes to a target nucleic acid and provides sequence information), a reactive group attached to a substrate (solid support), and a randomized Positions (ie, some of the four bases are found at these positions). Preferred probes have the sequence 5 ′-(T) _6- (N) _3- (B) ₅ (where T = thymine (attached to the solid support), N = A, C, G or T (random) B = 5 information providing positions (information providing portions) of the probe). In a most preferred embodiment, the probe is bound to a support and the spacer moiety is found 5 'to (N) ₃ at the end of the probe or within the probe. The spacer may be composed of atoms capable of forming at least two covalent bonds, such as carbon, silicon, oxygen, sulfur, and phosphorus, or may be a sugar-phosphate group, an amino acid, a peptide, a nucleoside, a nucleotide, a sugar, a carbohydrate, or an aromatic ring. , Hydrocarbon rings, linear and branched hydrocarbons and the like, which can form at least two covalent bonds.

Example 7 Production of Nucleic Acid Fragments Nucleic acids to be sequenced include cDNA, genomic DNA, chromosomal DNA, finely truncated chromosomal bands, cosmids or YAC inserts, and RNA including mRNA without any amplification steps. Obtained from any suitable source. For example, Sambrook et al. (1989) report three protocols for the isolation of high molecular weight DNA from mammalian cells (p. 9.14-9.23).

[0137] Target nucleic acid fragments can be produced as clones in M13, plasmid or lambda vectors, and / or can be produced directly from genomic DNA or cDNA by PCR or other amplification methods. The sample may be prepared or dispensed in a multiwell plate. About 100-
A 1000 ng DNA sample is prepared in a final volume of 2-500 ml. The target nucleic acid produced by PCR can be directly applied to a format 1 SBH substrate without purification. Once the target nucleic acid is immobilized on the substrate, the substrate can be washed or can be directly annealed with the probe.

The nucleic acid is then fragmented by any method known to those of skill in the art, including, for example, the use of restriction enzymes as described in Sambrook et al. (1989) 9.24-9.28, sonication and NaOH treatment. right.

The low pressure cleavage reported by Schriefer et al. (1990), which is incorporated herein by reference, is also suitable. In this method, a DNA sample is passed through a small French pressure cell at various pressures, from low to medium. Control of low to medium pressure applied to the cell by the lever device is possible. The results of these studies indicate that low pressure cleavage is a useful alternative to ultrasonic and enzymatic DNA fragmentation methods.

One particularly suitable method for DNA fragmentation would be to use the two base recognition endonuclease, CviJI, reported by Fitzgerald et al. (1992). The authors reported an approach for fragmenting DNA to specific sizes that would be suitable for rapid fragmentation and shotgun cloning and sequencing. We believe that this will also be particularly useful for generating random but relatively small DNA fragments for use in the present sequencing technology.

[0141] The restriction endonuclease CviJI usually cleaves the recognition sequence PuGCPy between G and C, leaving a blunt end. Typical reaction conditions that alter the specificity of this enzyme (CviJI **) give a quasi-random distribution of DNA fragments, forming the small molecule pUC19 (2688 base pairs). Fitzgerald et al. (1992) reported that the size-fractionated pUC19 C
Using viJI ** digestion, the randomness of this fragmentation method was assessed quantitatively and ligated directly to Lac Z minus M13 cloning vector without end repair. Sequence analysis of the 76 clones showed that CviJI ** restricted pyGCPy and PuGCPu in addition to the PuGCPy site, and that new sequence data accumulated at a rate consistent with random fragmentation.

As reported in the literature, the advantage of this approach over sonication and agarose gel fractionation is that it requires less DNA (0 to 2-5 μg).
.2-0.5 μg); and low number of steps (no need for pre-ligation, end repair, chemical extraction, or agarose gel electrophoresis and elution). These advantages have also been proposed to be useful when preparing DNA for sequencing by Format 3.

[0143] In a preferred embodiment, "fragments" of the nucleic acid sample are prepared such that they cannot be ligated together. A pool of such fragments is obtained by treating fragmented nucleic acids obtained by enzymatic digestion or physical cleavage with a phosphatase (eg, calf intestinal phosphatase). Alternatively, a non-ligated fragment of the sample nucleic acid is a random primer (eg, N ₅ -N ₉ , where N = A, G , T,
Or C). This will result in a DNA fragment that has a sequence complementary to the target nucleic acid, is terminated at a dideoxy residue, and cannot be ligated to other fragments.

Regardless of the method by which the nucleic acid fragment is obtained or produced, it is important to denature the DNA to provide a single-stranded fragment that can be used for hybridization. This is achieved by incubating the DNA solution at 89-90 ° C for 2-5 minutes. The solution is then rapidly cooled to 2 ° C. to prevent regeneration of the DNA fragment before contacting the chip.

Example 8 Production of DNA Array An array can be produced by spotting a DNA sample on a support such as a nylon membrane. Spotting is performed using an array of metal pins (corresponding to the array of wells in a microtiter plate) to a nylon membrane with approximately 20 nl of D.
The transfer of the NA solution can be repeated. Offset printing achieves a higher dot density than the well density. 1-25 dots, depending on the type of label used, can be accommodated in 1 mm ^2. By avoiding spotting in some preselected number of columns and rows, independent subsets (sub-arrays) can be formed. The samples in one subarray can be DNA segments of the same genome (or the same gene) from different individuals, or can be different, overlapping genomic clones. In one example, the selected gene segments can be amplified from 64 patients. For each patient, the amplified gene segments are in one 96-well plate (all 96 wells contain the same sample). Plates are manufactured for each of the 64 patients. By using a 96-pin device, all samples can be spotted on one 8x12 cm membrane. The subarray may contain 64 samples, one from each patient. If the 96 subarrays are identical, the dot spacing (span) can be 1 mm ² and the spacing between subarrays can be 1 mm.

Another approach is, for example, a physical grid such as a plastic grid molded over the membrane (the grid is similar to the type of membrane applied to the bottom of a multiwell plate), or a hydrophobic strip. Using a membrane or plate (available from NUNC, Naperville, Illinois) that can be divided by The fixed physical spacer is a flat phosphor-storage-screen (phospho
It is not preferred for imaging by exposure to r-storage screen) or X-ray film.

Example 9 Hybridization and Evaluation Process Labeled probes may be mixed with hybridization buffer and pipetted into a subarray (preferably by a multichannel pipette). To prevent mixing of the probes between the subarrays (when no hydrophobic bands or physical barriers are marked on the membrane), a corresponding grid of plastic, metal or ceramic is crimped to secure the membrane to the membrane. Is also good. Further, the volume of the buffer may be reduced to about 1 ml or less per 1 mm ² . The probe concentrations and hybridization conditions used may be as described above, but the wash buffer is quickly poured over a number of subarrays to facilitate rapid dilution of the probes and significant cross-hybridization. prevent. For similar reasons, the lowest concentration of probe may be used and the hybridization time extended to the longest at practical levels. For DNA detection and sequencing, knowledge of the "normal" sequence can enhance the signal using the phenomenon of continuous stacking interactions. In addition to the labeled probe, another unlabeled probe that continuously hybridizes with the labeled probe may be added to the hybridization reaction. The amount of hybrid may increase several times. Probes may be linked by ligation. This method can be important for breaking down DNA regions that form "compression".

In the case of radiolabeled probes, preferably fluorescent storage
ge) The technique makes it possible to obtain an image of the filter. Fluorescent labeling can be evaluated with a CCD camera, confocal microscope, etc. The raw signal is normalized based on the target amount for each dot in order to accurately score and summarize the data in different hybridization experiments. Differences in the amount of target DNA from dot to dot can be corrected by dividing the signal for each probe by the average signal of all probes evaluated on one dot. The normalized signal is scored (usually 1-100) to compare data from different experiments. Also, several control DNAs in each subarray can be used to determine the average background signal of a perfectly matched target-free sample. For a sample obtained by diploid (ploid) evaluation, a heterozygote in the sample can be recognized using a homozygous control.

Example 10 Hybridization with Oligonucleotides Oligonucleotides were purchased from Genosys Inc., Houston, Texas, or prepared on an Applied Biosystems 381A DNA synthesizer. Most of the probes used were not purified by HPLC or gel electrophoresis. For example, the probe is an M13 clone containing a 921 bp EcoRI-BglII human B1-interferon fragment in interferon [Ohno and Taniguchi, Proc. Natl. Acad. Sci. 74: 4270-4274.
(1981)], as well as a single target that is completely complementary, as well as at least one target that has a terminal base mismatch in the M13 vector itself.

End-labeling of oligonucleotides was performed as described [Maniatis et al., Molec.
ular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Cold Sp
ring Harbor, New York (1982)], T4-polynucleotide kinase (5 units, Amers
ham), γ ^{32 P-} ATP (3.3 pM, 10 mCi Amersham 3000 Ci / mM), and oligonucleotides (4 pM, 10 ng) in 10 ml. Specific activity of the probe is 2.5 ^-5 × 10 ⁹ cpm / nM
Met.

Single-stranded DNA (2-4 ml in a 0.5 NaOH, 1.5 M NaCl solution) was wetted with Gene Sc
Spotted on reen membrane. The filter was neutralized with 0.05 M Na ₂ HPO ₄ (pH 6.5), baked in a furnace at 80 ° C. for 60 minutes, and irradiated with ultraviolet light for 1 minute. The filters were then incubated in a hybridization solution (0.5 M Na ₂ HPO ₄ (pH 7.2), 7% sodium lauroyl sarcosine) for 5 minutes at room temperature, and the surface of a plastic Petri dish was removed. Placed. One drop of a hybridization solution (10 ml, 0.5 M Na ₂ HPO ₄ (pH 7.2), 7% sodium lauroyl sarcosine) containing ₄ nM concentration of ³² P-end labeled oligomer probe, 1 to 6 dots per filter The solution was dropped, covered with a square piece of polyethylene (about 1 × 1 cm), and incubated at a predetermined temperature in a humid chamber for 3 hours. Put the filter in 6X SSC washing solution at 0 ° C for 3 times.
The hybridization was stopped by removing the non-hybridized probe for 5 minutes. The filters were dried or further washed at the given times and temperatures and autoradiographed. For discrimination measurement, autoradiography [fluorescence imager (Molecular Dynamics, Sunnyva
le, California) can be used] and the dots were cut out of the dry filter afterwards, placed in a liquid scintillation cocktail, and counted. The uncorrected cpm ratio of IF dot and M13 dot is D.

The conditions reported herein are based on the ability to hybridize with very short oligonucleotides and to be complementary to the target nucleic acid, so that binding and mismatched oligonucleotides bind. It is like ensuring the distinction. Influence the effective detection of hybridization of short specific sequences based on the degree of discrimination (D) between perfectly complementary targets in a hybrid and incompletely complementary targets with a single mismatch. The influencing factors are defined. In experimental testing, dot blot hybridization of 28 probes, 6-8 nucleotides in length, to two M13 clones or model oligonucleotides bound to a membrane filter was achieved. The principles that guide the experimental process are described below.

Under probe-over conditions, oligonucleotide hybridization to a membrane-bound target nucleic acid (a few nucleotides longer than the probe) is a pseudo-primary reaction with respect to target concentration. This reaction: is defined by ^{_{S t / S 0 = e -kh}} [OP] t. Wherein, S _t and S _o are the target sequence concentration that put each time t and t _0. (OP) is the probe concentration and t is the temperature. Rate constant k _h hybridization is slightly increased in the range of 0 ℃ ~30 ℃ (Porschke and Eigen, J.Mol.Biol 62:.. 361 (1971); Craig et al., J. Mol Biol. 62 : 383 (1971)). Hybrid dissolution is a first order reaction for the hybrid concentration (replaced by mass due to membrane-bound state): H _t / H ₀ = e ^−kmt

[0154] In the formula, H _t and H ₀ is a hybrid concentration at each time t and t _0, k _m is dependent on temperature and salt concentration, which is the rate constant of the hybrid dissolution [described by Ikuta et al., Nucl. Acids Res. 15 : 797 (1987); Porsclike and Eigen, J. Mol.
Biol. 62 : 361 (1971); Craig et al., J. Mol. Biol. 62 : 303 (1971)]. During hybridization, the strand association process, reverse reactions, lysis reactions, or strand dissociation reactions also occur.
That is, the amount of hybrid formed in time results from forward and reverse reactions. Increasing the probe concentration and / or decreasing the temperature can direct the equilibrium to hybrid formation. However, in the wash cycle in large volumes of buffer, the lysis reaction is dominant and the reverse reaction hybridization is not significant due to the absence of the probe. This analysis shows that operable short oligonucleotide hybridization (SOH) condition requirements can be varied with probe concentration or temperature.

D or identification is defined by equation 4: D = H_p(t_w) / H_i(t_w) H_p(t_w) And H_i(t_w) Indicates the same amount of perfect complementary duplex and incomplete phase, respectively.
For complementary duplexes, the washing time t_wIs the amount of hybrid remaining after The discrimination D 1 varies at a given temperature according to a length of cleaning time of 10; _i The maximum value is reached when = B.

Background B represents the lowest hybridization signal detectable in the system. Since the more reduced H _i may not be inspected, D is increased and washed in succession. washed more than t _w is only reduces the H _p for B, manifested as a decrease in D. Optimum cleaning time t _w incomplete hybrid, from Equation 3 and Equation _{5: t w = -In (B} / H i (t 0)) is / K _mi.

Since H _{p is} also cleaned by the same t _w , combining the equations yields the following optimal discriminant function: D = e ^{In (B / Hi (t0)) km, p / km, i} XH _p The change in D as a function of (t ₀ ) / BT is important for selecting the optimal cleaning temperature. This is obtained by substituting the ^Arhenius equation: K− = Ae− ^{Ea / RT} into the above equation, and ^obtaining the final equation: D = H _p ((t ₀ ) / BX (B / H _i (t ₀ )) ) Obtain ^{(Ap / Ai) e (Ea, i-Ea, p) / RT} .

Where B is lower than H _i (t ₀ ).

The activation energy E _{a, p of} the perfect hybrid and the activation energy E _{a, i} of the incomplete hybrid are either equal or E _{a, i} <E _{a, p} , and therefore in each case D is independent of temperature or decreases with increasing temperature. This result indicates that it is inappropriate to examine stringent temperature conditions for good discrimination in SOH. Washing at lower temperatures gives equivalent or better discrimination, but the washing time increases exponentially with decreasing temperature. H _p (
If H _i (t ₀ ) increases with respect to t ₀ ), the discrimination decreases more strongly with T.

D at low temperatures is more dependent on the H _p (t ₀ ) / B ratio than on the H _p (t ₀ ) / H _i (t ₀ ) ratio. This result is independent of the identity obtained in this step indicates that it is better to obtain a sufficient amount of H _p in the hybridization. Better identification can be obtained by washing. Because the higher the amount of complete hybrid, the longer it takes for differential lysis to take effect. Similarly, higher amounts of target nucleic acid will provide the necessary discrimination even if the difference between K _{m, p} and K _{m, i} is small.

When estimating situations that are more complex than those covered by this simple model, in the case of hybridization of probes with many end mismatches at a given nucleic acid target, low temperatures are required to obtain discrimination. Cleaning becomes more important.

Using the theoretical principles described as experimental guidelines, reliable hybridization is obtained with probes 6-8 nucleotides in length. All experiments were performed on floating plastic sheets with the hybridization solution filmed on filters. This method minimizes the amount of probe, thus reducing the cost of labeling in dot blot hybridizations. Using high concentrations of sodium lauroyl sarcosine instead of sodium lauroyl sulphate in the phosphate hybridization buffer can reduce the reaction from room temperature to 12 ° C. Similarly, 4 ⁻⁶ × SSC, 10% sodium lauroyl sarcosine buffer allows hybridization at a low temperature of 2 ° C. Detergents in these buffers are for obtaining acceptable background with labeled probes at concentrations up to 40 nM. The initial properties of the thermal stability of short oligonucleotide hybrids were determined by using a prototype 8-mer with a G + C content of 50% (
That is, a probe of the sequence TGCTCATG). Theoretically, this probe is one of the more unstable 8-mers. The transition enthalpy of this probe is close to that of the more stable 7-mer, even to a 6 nucleotide long probe (Bre
Sslauer et al., Proc. Natl. Acad. Sci. USA 83 : 3746 (1986)). The parameter T _d (1
The temperature at which 50% of the hybrid dissolves per unit time per minute) is 18 ° C. The result is a hybrid of the T _d of 8bp indicates that lower 15 ℃ than T _d of the double helix of 11 bp [Wallace et al., Nucleic Acids Res 6:. 3543 (1979)].

In addition to experiments with model oligonucleotides, the M13 vector was selected as a system for proving the performance of short oligonucleotide hybridizations. The primary objective was to demonstrate useful end-pair mismatch discrimination using targets similar to those used in various applications of the method of the present invention. Oligonucleotide probes for the M13 model were selected such that the M13 vector itself contained terminal mismatched bases. Vector IF (M13 recombinant containing the 921 bp human interferon gene insert) carries a single perfect-matched target. Thus, IF has the same or more mismatched targets than the M13 vector itself.

Using low temperature conditions and dot blots, a sufficient difference in hybridization signal can be obtained between dots containing perfectly matched and mismatched targets and dots containing only mismatched targets. It is. This was also true for 6-mer oligonucleotides, as well as 7-mer and 8-mer oligonucleotides that hybridized to large IF-M13 paired nucleic acids.

[0165] The hybridization signal depends on the amount of target on the filter that can react with the probe. Controls are needed to show that the difference in signature strength between the two dots does not reflect the difference in the amount of nucleic acid. Hybridization with a probe having the same number and type of targets in both IF and M13 indicates that the same amount of DNA is present in the dot. Since the efficiency of hybridization increases with the length of the hybrid, the signal of a double helix with six nucleotides was best detected with a high mass oligonucleotide target bound to the filter. Oligonucleotide target molecules with lower molecular weights can bind more to a given surface area than nucleic acids of larger molecules that function as targets.

To determine the sensitivity of detection on unpurified DNA, various amounts of phage supernatants were spotted on filters and hybridized with ³² P-labeled 8-mers. Detectable signals were obtained from a minimum of 50 million crude phages containing only 0.5 ng of DNA, demonstrating that the short oligonucleotide hybridization method was sensitive enough. On the practical side, it adds that the reaction time is short.

As described in the theory section above, the equilibrium yield of the hybrid depends on the probe concentration and / or the reaction temperature. For example, with 4 nM 8-mer at 13 ° C., the signal level of the target is 3 times lower than the same amount of target signal at 40 nM probe concentration, and 4.5 times lower when the hybridization temperature is raised to 25 ° C. Become.

Washing at low temperatures to obtain maximum discrimination has been demonstrated. To clearly visualize the phenomenon, 50 times more DNA was added to the M13 dot compared to the IF dot, and hybridization with a vector-specific probe was used. In this way, the signal after the hybridization step with the actual probe was stronger in the case of mismatch than in the case of pairing. The _Hp / _Hi ratio was 1: 4. Washing for an extended period of time at 7 ° C. provided a reversal of signal intensity at a 2: 1 ratio without losing large amounts of complete hybrid. Conversely, it was not possible to obtain any discrimination at 25 ° C.
This is because a two minute wash will already bring the paired target signal down to background values, while the signal from the mismatched hybrid will remain detectable. The loss of discrimination at 13 ° C. is less significant than at 7 ° C., but is clearly visible. Considering the 90 min time point at 7 ° C. and the 15 min time point at 13 ° C., if the mismatched hybrid signal is close to the background value representing the optimal wash time in each condition, it will be more than 7 ° C. It is clear that the amount is several times higher at ° C. To further illustrate this, the change in discrimination over time due to washing of the same amount of starting hybrid at two temperatures indicates a higher maximum D at lower temperatures. From these results, it is possible to confirm the tendency of the change of D according to the temperature and the ratio of the amounts of the two hybrids at the start of the washing step.

To demonstrate the general utility of short oligonucleotide hybridization conditions, in our simplified M13 system, four 7-mers, ten 8-mers, and 14 up to 12 nucleotides in length were used. Hybridization of the probe was observed.
These represent the two extremes of GC content, 9-merGTTTTTTAA and 8-merGGCAGGC.
Including G. GC content and sequence are expected to affect the stability of short hybrids [Bresslauer et al., Proc. Natl. Acad. Sci. USA 83 : 3746 (1986)].
Low temperature, short oligonucleotide conditions could be applied to all probes tested to obtain sufficient discrimination. The best discriminating value obtained with a 13 nucleotide long probe is 20, and a several-fold reduction due to sequence changes is readily acceptable.

The M13 system has the advantage of indicating the effect of target DNA complexity on the level of discrimination. For the two 8-mers that had 0 or 5 mismatched targets and differed only by one GC pair, the discrimination observed was 18.3 and 1.7, respectively.

To demonstrate the usefulness of this method, three 8-nucleotide long probes were used
Tests were performed on a set of 51 plasmid DNA dots generated from a library of script vectors. One of the probes was present and specific for the Bluescript vector, but was absent for M13. The other two probes had targets that were known sequence inserts. In this system, hybridization with negative or positive control DNA for each probe can be performed. This probe sequence (CTCCCTTT) also had a complementary target in the interferon insert. Hybridization was sequence specific because the M13 dot was negative and the interferon insert in M13 or Bluescript was positive. Similarly, the probe detects the target sequence in only one of the 51 inserts, the control does not detect any target sequence in the test insert, and the hybridization is successful if the appropriate target is present in the clone. It was confirmed that it would have occurred.

Thermal stability curves for very short oligonucleotide hybrids of 6-8 nucleotides in length are at least 15 ° C. lower than hybrids of 11-12 nucleotides in length [FIG. 1 and Wallace et al., Nucleic Acids Res. . 6 : 3543-3557 (1979)]
. However, performing the hybridization reaction at low temperatures and with a very practical concentration of 0.4-40 nM oligonucleotide probes allows detection of complementary sequences in known or unknown nucleic acid targets. To completely determine the unknown nucleic acid sequence, a complete set containing 65,535 8-mer probes can be used. A sufficient amount of nucleic acid for this purpose can be obtained from any convenient biological sample (eg, microliters of M13 culture, plasmid preparation or single colony bacteria from 10 ml of bacterial culture, less than 1 ml). In standard PCR reactions).

Excellent discrimination is obtained from oligonucleotides as short as 6 to 10 nucleotides in length. The relative decrease in hybrid stability due to single-terminal mismatches is greater than for longer probes. The results obtained from 8-merTGCTCATG support this conclusion. In experiments, hybridization to targets with G / T end mismatches, targets with this type of mismatch, is the most stable of all other species of oligonucleotides. The resulting discrimination is the same or greater than the internal G / T mismatch in a 19 base pair duplex (Ikuta et al., Nucl. Acids res. 15: 797 (1987)). Exploiting these discriminating properties using the described hybridization conditions for short oligonucleotides allows very accurate determination of the oligonucleotide target. For the ease of detecting the distinction between perfect and incomplete hybrids, the problem that exists with using very short oligonucleotides is to prepare sufficient quantities of hybrids. In fact, the need to identify the H _p and H _i is aided by lowering to increase the amount of DNA and / or probe concentration in the dot or the hybridization temperature. However, high probe concentrations usually increase background. In addition, there is a practical limit on the amount of target nucleic acid. This problem was solved by increasing the concentration of the detergent sarkosyl, which provided an effective background with 4 nM probe. Further improvements can be made by using competitors to prevent non-specific binding of the probe to the filter, or by changing the hybridization support material. Further, the probe (e.g., a number of 7-mer and major 6-mer) having the E _a of less than 45 kcal / mol for the direction of modified oligonucleotides, unmodified equivalents resulting stable hybrid than [Asseline et al., Proc. Natl. Acad. Sci. 81 : 3297 (1984)]. The hybridization conditions described in the present invention for hybridization with short oligonucleotides utilizing low temperatures provide better discrimination for all sequences and duplex hybrid inputs. The only trade-off for obtaining uniformity of hybridization conditions for different sequences is that the wash time can be increased from a few minutes up to 24 hours depending on the sequence. The cleaning time is
It can be further shortened by lowering the salt concentration.

Although short oligonucleotide hybridization is excellent in discriminating one paired hybrid against a mismatched hybrid, signals from mismatched hybrids are also present. Many of these mismatched hybrids are due to terminal mismatches. This can limit the size of the insert that can be effectively tested with a particular length of probe.

The impact of sequence complexity on identification is not negligible. But the impact of complexity is
Hybridization with short oligonucleotides is more significant when defining sequence information for specific and non-random sequences and can be overcome by using the ratio of (appropriate probe length) to (target length) . The length ratios are chosen based on statistics to eliminate the possibility of generating specific sequences with some end mismatches that either eliminate the identification or incorrectly reverse the identification. The results show that for target nucleic acid inserts smaller than 0.6, 2.5, and 10 kb,
, And the use of oligonucleotides 8 nucleotides in length is suggested.

Example 11 DNA Sequencing An array of multiple subarrays allows for efficient sequencing of small sets of samples arranged in repeated subarrays. For example, 64 samples are arranged on an 8 × 8 mm sub-array, and a 16 × 24 sub-array is 15 × 23 cm.
May be repeated at 1 mm width intervals on the film. Several replica films may be made.
For example, a probe from one universal set consisting of 3072 7-mers,
It can be divided into 32 96-well plates and labeled with kinase. Four membranes may be processed in parallel during one cycle of hybridization. 384 probes may be scored on each membrane. All probes may be scored in two cycles of hybridization. The hybridization intensity is scored and the sequences can be assembled as follows.

If the sub-array of single or multiple samples contains several unknowns, especially if similar samples are used, judicious selection of the probes based on the results of the previously scored probes, Fewer probes are sufficient. For example, if the probe AAAAAAA is not positive, it is unlikely that any of the eight overlapping probes is positive. If the AAAAAAA is positive, usually the two probes are positive. In this case, the sequencing process, which identifies one of the most likely hypotheses about the order of the anchors and the size and type of gap between each anchor, first hybridizes a subset of probes with minimal overlap. And then selecting the probe after defining the positive anchor.
In this second step, if each probe was selected to be positive in only one DNA sample different from the sample expected to be positive for other probes from the pool, A pool of 10 probes can be used.

This subarray method allows efficient implementation of probe competition (overlapping probes) or probe cooperation (continuous probe stacking) to resolve branching problems. After hybridizing the universal set of probes, a subfragment (SF) of the candidate sequence is determined by a sequence assembly program. Additional information must be provided (from overlapping sequences of DNA fragments, analogous sequences, single pass gel sequences, and other hybridization or restriction map data) for further assembly of the SF. Primers for single-pass gel sequencing through the branch point are identified from SBH sequence information or from known vector sequences (eg, flanking sequences at vector insertion sites), and standard Sanger sequencing reactions are performed on sample DNA. Do it for The sequence obtained from this single pass gel sequencing is compared to the Sf read into or read from the branch point to identify the order of the Sf. In addition, single pass gel sequencing can be combined with SBH to newly sequence or resequence nucleic acids.

Sf can also be assembled using competitive hybridization and continuous stacking interactions. These methods have limited commercial value for sequencing large samples because SBH applies labeled probes to samples immobilized on the array when using a homogeneous array. Fortunately, analysis of small samples using replica subarrays allows for efficient implementation of both methods. On each replica subarray, using a pool of probes,
One branch point can be tested on one or more DNA samples in the same way as elucidating mutant sequences in different samples spotted in the same subarray (see above).

Each of the 64 samples described in this example has 100 bifurcation points, and if 8 samples are analyzed in parallel on each subarray, at least 800 subarray probing will resolve all bifurcation points. . This means that an additional 800 probings (25%) are used for 3072 base probings. More preferably, two probings are used at one branch point. If the subarray is smaller, less probing is used. For example, if the subarray is 16
When consisting of 200 samples, 200 additional probings are scored (6%). 7-
Assembling a fragment of about 1000 bp by probing about 4000 using a mer probe (N _1-2 B ₇ N _1-2 ) and using competitive or collaborative branch point resolution methods or both. Can be. In addition, 4 kb or longer fragments can be assembled with 12000 probing using an 8-mer probe (NB ₈ N). Gapped probe, eg NB ₄ NB ₃ N or NB ₄ NB ₄ N
Can be used to reduce the number of branch points.

[0181] ra DNA analysis and labeled probe by binding to a temporary subarray of Example 12 Probe Igeshon

Oligonucleotide probes having an informative length of 4 to 40 bases are synthesized by standard chemistry and stored in test tubes or multiwell plates. A particular probe set consisting of 1 to 10,000 probes is aligned by sedimentation on a separate support or a different portion of a larger support, or by in situ synthesis. In the latter case, the portions or subarrays may be separated by a physical or hydrophobic barrier. This probe array may be prepared by in situ synthesis. Hybridize the appropriately sized sample DNA in one or more specific arrays. Many of the samples may be computerized as pools in the same subarray, or each with a different subarray in one support. Simultaneously with or following the sample, a single labeled probe, or one pool of labeled probes, is added to each subarray. When the bound and labeled probes are hybridized back-to-back to complementary targets in the sample DNA, they are ligated. The occurrence of ligation is measured by detecting the label from the probe.

This procedure differs from the described DNA analysis method in that the DNA sample is not permanently bound to the support. Temporary binding is provided by a probe immobilized on a support. In this case, there is no need for a target DNA array method. In addition, ligation allows detection of longer oligonucleotide sequences by conjugating a short labeled probe with a short immobilized probe.

This method has several unique features. Basically, the temporary binding of the target allows its reuse. After ligation has occurred, the target may be released and the label remains covalently attached to the support. This feature allows cycling of the target and production of a signal that can be detected with small amounts of target. Under optimal conditions,
There is no need to amplify the target. For example, a natural source of a DNA sample may be used directly for diagnostic and sequencing purposes. The target may be released by cycling the temperature between efficient hybridization and efficient duplex melting. The temperature and concentration of the components may be defined to have a balance (approximately 50: 50%) between the free target and the target in the hybrid. In this case, the linked product is produced continuously. For different purposes, the optimal equilibrium ratio will be different.

An electric field may be used to enhance target utilization. First, horizontal field pulsing within each sub-array may be used to target faster. At this stage, the equilibrium moves to hybrid formation,
Unlabeled probes can be used. After going through the target selection step, a suitable wash may be performed, which may be facilitated by restricting the movement of the sample with a vertical electric field. Several cycles of discriminating hybrid melting, harvesting the target by hybridization, ligation and removal of unused target may be introduced to increase specificity. In the next step, labeled probes may be added and vertical electrical pulses may be used. The temperature is increased to obtain the optimal free to hybridized target ratio. The vertical electric field prevents diffusion of the sorted target.

The set of immobilized probe subarrays and labeled probes (particularly designed or selected from common probe sets) may be variously arranged to allow for efficient and flexible sequencing and diagnostic methods. . For example, to partially or completely sequence short fragments (about 100-500 bp) of the bacterial genome, use a small array of probes (5-30 bases long) designed on bases of known sequence. May be. Examining a separate pool of 10 labeled probes per subarray, ie, every 10 subarrays with 10 probes each, can identify 200 bases, assuming that 2 bases linked by ligation are evaluated. Under conditions that identify mismatches through hybrids, the probe may be replaced with one or more bases to cover a longer target with the same number of probes.
The use of long probes allows direct interrogation without amplification or isolation from the remaining DNA in the sample. Also, several targets may be analyzed (screened) in one sample at the same time. If the results obtained indicate the occurrence of a mutation (or pathogen), additional pools of probes may be used to detect the type of mutation or subtype of pathogen. This is a feature of this method and is cost-effective in prophylactic diagnosis where only a few patients are considered to have the infection or mutation.

In the methods described in the embodiments, various detection methods may be used. For example,
Large molecules or particles that can be detected by methods that utilize radiolabels, fluorescent labels, enzymes or antibodies (chemiluminescence), light scattering or light interference.

Example 13 Target Sequencing Using 8-mers and 9-mers The data obtained from hybridization of 8-mer and 9-mer oligonucleotides shows that sequencing by hybridization is very accurate. It is shown that. In this experiment, the 8-mer and 9-mer of adjacent overlapping components
The known sequence was used to deduce the -mer oligonucleotide.

In addition to perfectly matched oligonucleotides, mismatched oligonucleotides, mismatched oligonucleotides in which internal or terminal mismatches occur in the duplex formed by the oligonucleotide, and target Was examined. In this analysis, the lowest actual temperature was used to maximize hybridization formation. Washing was performed at the same or lower temperature as above to ensure maximum discrimination using the maximum dissociation rate of mismatched paired oligonucleotide / target hybridization. Absolute hybridization yields are shown as sequence dependent, but the above conditions are shown as applicable to all sequences.

Since the mismatch that can be assumed to have minimal instability is a single-ended mismatch, testing for sequencing by hybridization means that a perfectly matched oligonucleotide / target duplex The ability to distinguish a helix from a mismatched oligonucleotide / target duplex at the ends.

For the 102 to 105 hybridizing oligonucleotides in the dot blot format, the value indicating discrimination was greater than 2, which allowed for accurate sequence generation. The system also allows for the analysis of the effect of sequences on hybridization formation as well as hybridization instability.

Using data obtained by hybridization of 105 oligonucleotide probes of known sequence to a target nucleic acid, for example, PCR of a 100 bp target sequence
Generated 100 base pairs of the known portion of the human interferon gene prepared in. The oligonucleotide probes used included 72 8-mer and 21 9-mer oligonucleotides whose sequences were completely complementary to the target. 93 probe sets provided a contiguous overlapping frame of target sequence shifted by one or two bases.

To assess the effect of the mismatch, the hybridization was examined for 12 additional probes containing at least one mismatch when hybridized to a 100 bp test target sequence. We also tested the hybridization of 12 probes with targets that end mismatched with the other four selected control nucleic acid sequences,
The 12 oligonucleotides formed were perfectly paired with a duplex hybrid with four control DNAs. Thus, internal mismatches, terminal mismatches, and hybridization of perfectly matched double helix pairs of oligonucleotide and target were evaluated for each of the oligonucleotides used in the test. The effect of absolute DNA target concentration on hybridization using the 8-mer and 9-mer oligonucleotides of the test was determined by the use of different oligonucleotide probes to a single occurring non-target site in co-amplified plasmid DNA. Detecting hybridization,
It was determined by defining the target DNA concentration.

The results of this test indicated that all oligonucleotides containing complementary sequences that perfectly matched the target, or control DNA, hybridized more strongly than the oligonucleotides with the mismatch. To reach this conclusion, the Hp and D values for each probe were examined. Hp defines the amount of hybrid duplex formed between the test target and the oligonucleotide probe. 105 between 0 and 10
Exposure to the hybridization obtained for probe No. 6 revealed that 68.5% of the 105 probes had an Hp greater than 2.

1) Dots containing one perfectly matched double helix formed between the test oligonucleotide and the target or control nucleic acid; and 2) the same oligonucleotide and different sites within the target or control nucleic acid. The discrimination (D) value was obtained by defining the ratio of the signal strength between the dot containing one mispaired double helix formed between it and the dot. Variations in D-values can be attributed to 1) a perturbation of the effectiveness of hybridization, which allows visualization of the signal from the background, or 2) the type of mismatch found between the test oligonucleotide and the target. Brought by either. The D value obtained in this test was 2 for 102 of the 105 oligonucleotide probes tested.
To 40. For a group consisting of 102 oligonucleotides as a whole,
When the D value was calculated, the average was 10.6.

The oligonucleotide / target duplexes showed terminal mismatch in 20 cases. In five of these, D was greater than 10. The large D value in this case is likely due to hybridization instability caused by other than the most stable (G / T and G / A) end mismatches. Other possibilities include errors in the sequence of either the oligonucleotide or the target.

The possibility of errors in the target of the low Hp probe was ruled out, as such errors would have affected the hybridization of each of the other eight overlapping oligonucleotides. The lack of apparent instability due to sequence mismatch for the other overlapping oligonucleotides indicates that the target sequence was correct. Possible errors in the oligonucleotide sequence were ruled out after re-examining the seven newly generated oligonucleotides. Only one of the seven oligonucleotides gave better D values. Low hybridization values can result from hybrid instability or inability to form hybrid duplexes. Inability to form hybrid duplexes can occur either from 1) self-complementarity of the chosen probe or 2) target / target self-hybridization. If the probe was self-complementary, oligonucleotide / oligonucleotide duplex formation may be more advantageous than oligonucleotide / target hybrid duplex formation. Similarly, if the targets were self-complementary, target / target binding could be advantageous or could form an internal palindrome. Upon evaluation of these possibilities, probe analysis revealed that the problematic probe did not hybridize to itself. In addition, examination of the contribution of target / target hybridization determined that one of the problematic oligonucleotide probes hybridized inefficiently with two different DNAs containing the same target. The low probability that two different DNAs have self-complementary regions for the same target sequence leads to the conclusion that target / target hybridization did not contribute to low hybridization formation. Thus, these results suggest that the instability of the hybrid, rather than the inability to hybridize, was responsible for the low hybridization observed for certain oligonucleotides. These results also suggest that low hybridization is due to the specific sequence of certain oligonucleotides.
Furthermore, these results suggest that the use of 8-mer and 9-mer oligonucleotides would provide reliable results for generating sequences.

These results indicate that using the described method, long sequences of any particular target nucleic acid will be produced with maximal and unique overlap of the constituent oligonucleotides. . Such sequencing methods are independent of their frequency or position and depend on the content of the individual constituent oligomers.

An array generated using the following algorithm has high accuracy. This algorithm allows for false positive signals from the hybridization dots, as indicated by the fact that the sequence generated from the 105 hybridization values, including four lower confidence values, was correct. Accuracy in sequencing by hybridization is less than "one or eight" of short oligonucleotide hybridizations.
all or none) "as well as differences in stability of the duplex between perfectly paired and mispaired duplexes. The ratio of duplex stability to paired and terminally mismatched duplexes increases with decreasing duplex length. In addition, the binding energy decreases as the duplex length decreases, resulting in lower hybridization efficiency. However, these results indicate that octamer hybridization averages factors that affect duplex helix stability,
As well as discrimination to produce a more accurate method of sequencing by hybridization. The results in other examples are 6,7,
Alternatively, oligonucleotides of eight nucleotides have been shown to be useful for generating reliable sequences for targets of 0.5 kb (hexamer), 2 kb (septamer), and 6 kb (octamer). The sequences of the long fragments may overlap to generate a complete genomic sequence.

Example 14 Analysis of Data Obtained An image file was analyzed by an image analysis program such as the DOTS program (Drmanac et al., 1993). The evaluation was performed by scaling with the objective function. From the signal distribution, determine the optimal threshold for converting the signal to +/- output. From the position of the detected sign,
The F + P nucleotide sequence from the fragment will be determined by combining the labeled probe corresponding to the labeling position with the known sequence of the immobilized probe. Thereafter, the complete nucleic acid sequence or sequence subfragment of the original molecule, such as the human chromosome, will be assembled from overlapping F + P sequences determined by computer deduction.

One option is to convert the hybridization signal (eg, score) to +/- output during the sequence assembly. In this case, the assembly will start from the F + P sequence with a very high score, for example the F + P sequence AAAAAATTTTTT. All four possible overlapping probes AAAAATTTTTA, AAAA
Comparing the scores of ATTTTTT, AAAAATTTTTTC and AAAAATTTTTTG, and three additional probes that differ at the beginning (TAAAAATTTTTT; CAAAAATTTTTT; GAAAAATTTTTT)
Three results are defined. That is, (i) only one of the four overlapping probes with the starting probe has a significantly positive score relative to the other six probes. In this case, the AAAAAATTTTTT sequence is extended to the right by one nucleotide; (ii) none of the probes except the starting probe have a significantly positive score.
Assembly stops, eg, the AAAAAATTTTT sequence comes to the end of the DNA molecule to be sequenced; (iii) there are two or more significantly positive scores among the overlapping probes and / or the other three probes. Assembly is stopped due to errors or branching (Drmanac et al., 1989).

In computer deduction methods, computer programs using existing algorithms will be used (eg, Pevzner, 1989; Drmanac et al., 1991; Labat & Dr.
manac et al., 1993; each of which is incorporated herein by reference).

If F (space 1) P, F (space 2) P, F (space 3) P or F (space 4) P is determined in addition to F + P, all Matching the datasets will correct possible errors and clarify instances where branching problems exist (eg, Drmanac et al., 1989; Bains et al., 1988; each of which is incorporated herein by reference).

Example 15 Following several examples describing the implementation of the sequencing method contemplated by the present inventors for performing sequencing by two-step hybridization , first, the entire chip was converted to 100 million bp (1 (A human chromosome in a book). Guidelines for performing hybridization are given in documents such as Drmanac et al. (1990); Khrapko et al. (1991); Broude et al. (1994). These references teach suitable hybridization temperature ranges, buffers and washing steps to use in the initial steps of Format 3 SBH.

The inventor has determined that hybridisation should be carried out at high salt concentrations at low temperatures (−2 to 5 ° C.) for up to several hours, especially because of the relatively low concentrations of target DNA that can be provided. Is intended. For that, an SSC buffer is used instead of a sodium phosphate buffer that precipitates at 10 ° C. (Drmanac et al., 1990). Washing need not be sufficient for the second step (a few minutes) and can be completely eliminated when using hybridization cycling for sequencing highly complex DNA samples. If the same buffer is used in the hybridization step and the washing step, the second hybridization step using the labeled probe can be continued.

After appropriate washing using a simple robotic device for each array (eg, an 8 × 8 mm array), one labeled probe (eg, a 6-mer) is added. This is performed 42 times using a 96-chip or 96-pin device. Again, a range of discriminatory conditions will be employed, as described in the scientific literature.

The inventor specifically intends to employ the following conditions. First, add the labeled probe and at low temperature (0-5 ° C) (due to the high concentration of added oligonucleotide)
After incubating for only a few minutes, raise the temperature to 3-10 ° C., depending on the F + P length, and add wash buffer. The washing buffer used at this time is compatible with any ligation reaction (for example, a salt concentration range of 100 mM). After the addition of ligase, the temperature is raised again to 15-37 ° C to allow for rapid ligation and further discrimination of perfect and mismatched hybrids.

As described in Pontius & Berg (1991, incorporated herein by reference):
The use of a cationic surfactant is also contemplated in Format 3 SBH. These authors describe the use of two simple cationic surfactants, dodecyl- and cetyltrimethylammonium bromide (DTAB and CTAB) in DNA recombination.

DTAB and CTAB are the quaternary amines tetramethylammonium bromide (
TMAB), in which one of the methyl groups has been replaced with either a 12-carbon alkyl group (DTAM) or a 16-carbon alkyl group (CTAB). TMAB is a bromide salt of tetramethylammonium ion, a reagent used in nucleic acid recombination experiments to reduce the GC content bias in melting temperature. DTAB and CTAB are structurally similar to sodium dodecyl sulfate (SDS), where the negatively charged sulfate group of the SDS is replaced by a positively charged quaternary amine. SDS is commonly used in hybridization buffers to reduce non-specific binding and inhibit nucleases, but does not significantly affect rebinding rates.

When using the ligation method, the enzyme will be added with the labeled probe or after an appropriate washing step to reduce background. Although not previously suggested for use in the SBH method, the ligase technique is well established in the field of molecular biology. For example, Hood and co-workers have described a ligase-mediated gene detection method (Landegren et al., 1988), which can be readily adapted for use in Format 3 SBH. Wu & Wallace also describes the use of bacteriophage T4 DNA ligase to ligate two adjacent short synthetic oligonucleotides. Their oligo ligation reaction was performed with 50 mM Tris
HCl pH 7.6, 10 mM MgCl ₂ , 1 mM ATP, 1 mM DTT and 5% PEG. Heat the ligation reaction to 100 ° C. for 5-10 minutes, then cool to 0 ° C. before T4 DN
A ligase (1 unit; Bethesda Research Laboratory) was added. Most ligation reactions were performed at 30 ° C. and terminated by heating at 100 ° C. for 5 minutes.

Thereafter, a final wash suitable for the differential detection of hybridized adjacent or ligated, length (F + P) oligonucleotides is performed. This washing step is performed for several minutes in water at 40-60 ° C. to wash away all unligated labeled probes and all other compounds to minimize background. Because of the covalently linked labeled oligonucleotide, detection is simplified (time and cold are not forced).

Depending on the label used, imaging of the chip is performed using different devices. In the case of radioactive labels, phosphor storage screen technology and scanner
PhosphorImager can be used (Molecular Dynamics, Sunnyvale, CA)
. Place the chips in the cassette and cover with a phosphor screen. After 1-4 hours of exposure, scan the screen and store the image file on the computer's hard disk. For detection of fluorescent labels, use a CCD camera and epifluorescence or confocal microscope. For chips fabricated directly on the pixels of a CCD camera, E
Detection can be performed as described by Gggers et al. (1994, incorporated herein by reference).

[0213] Charge-coupled device (CCD) detectors serve as active solid supports that quantitatively detect labeled target molecules in a probe-based assay and image their distribution. These devices use the unique features of microelectronics that are compatible with highly parallel assays, ultrasensitive detection, high throughput, integrated data acquisition and computation. Eggers et al. (1994) describe a CCD for use with a probe-based assay, such as the format 3 SBH of the present invention, which is within a few seconds due to the high sensitivity and direct binding employed. Enable quantitative evaluation.

The integrated CCD detection method allows for the detection of molecular binding events on the chip. The detector quickly creates a two-dimensional pattern that uniquely characterizes the sample. CC
In a specific operation of a molecular detector based on D, individual biological probes are
It may be fixed directly on the pixel of interest, or coupled to a disposable cover slip placed on the CCD surface. Sample molecules can be labeled with radioisotopes, chemiluminescent or fluorescent tags.

Exposure of a sample to a CCD-based probe array generates, in the case of format 3, photon or radioisotope decay products at pixel locations where the sample is bound to two complementary probes. Thereafter, when radiation from charged particles or labeled sample is incident on the CCD gate, electron-hole pairs are formed. Next, electrons are collected under the adjacent CCD gate and read sequentially by the display module. The number of photoelectrons generated at each pixel is directly proportional to the number of such nearby molecular binding events. As a result, molecular binding can be measured quantitatively (Eggers et al., 1994).

Placing the imaging array in close proximity to the sample increases the collection efficiency by at least 10 times over lens-based techniques as found in conventional CCD cameras. That is, the sample (emitter) is in close contact with the detector (imaging array), which eliminates conventional imaging optics such as lenses and mirrors.

If a radioisotope is attached to the target molecule as a reporter group, energetic particles are detected. And successful with several reporter group capable of generating a variety of energy of the particles is used with microfabrication detector, and the like ^{32 P, 33 P, 35 S} , 14 C and ¹²⁵ L. High energy particles, such as from ³² P, provide the highest molecular detection sensitivity, while low energy particles, such as from ³⁵ S, provide better resolution. Therefore, the radioisotope reporter can be selected according to requirements. Once a particular radioisotope label has been selected, detection performance can be estimated by calculating the signal-to-noise ratio (SNR), as described by Eggers et al. (1994).

Another luminescence detection method involves the use of a fluorescent or chemiluminescent reporter group attached to the target molecule. Fluorescent labels can be attached covalently or via interaction. Fluorescent dyes such as ethidium bromide have strong absorption bands in the UV (300-350
nm) and the main emission band is in the visible (500-650 nm) range, making it most suitable for the CCD device used. This is because the quantum efficiency is several orders of magnitude lower at the excitation wavelength than at the fluorescence signal wavelength.

From the perspective of detecting luminescence, the polysilicon CCD gate has a built-in capacity to filter the contribution of incident light in the UV range and is very sensitive to visible luminescence generated by the fluorescent reporter group. Good. Such a large unique discrimination for UV excitation is due to the large SNR in the CCD proposed by Eggers et al. (1994).
(Greater than 100).

For immobilization of the probe to the detector, a hybridization matrix can be made on an inexpensive SiO 2 wafer, which is then placed on the surface of the CCD after hybridization and drying. This format is economically efficient. This is because DNA hybridization is performed on inexpensive disposable SiO2 wafers, thus allowing the reuse of relatively expensive CCD detectors. Alternatively, a probe matrix may be prepared by immobilizing a probe directly on a CCD.

To immobilize the probe on the SiO ₂ coating, a uniform epoxide layer is attached to the film surface using epoxy-silane reagents and standard SiO ₂ modification chemistry. Next, the amine-modified oligonucleotide probes are bound to the SiO ₂ surface by secondary amines formation of epoxide ring. The resulting linkage provides 17 rotatable linkages that separate the three bases of the oligonucleotide from the SiO ₂ surface. The reaction is performed in 0.1 M KOH and incubated at 37 ° C. for 6 hours to ensure complete amine deprotonation and to minimize the formation of secondary structure during binding.

Generally, in format 3 SBH, the signal is evaluated at each of a vast number of points. It is not necessary to hybridize all arrays, for example 4000 5x5 mm, at a time, and a smaller number of arrays can be used.

[0223] Cycling hybridization is one method that can enhance the hybridization signal. In one cycle, most of the immobilized probe will hybridize to a DNA fragment having a tail sequence that is non-complementary to the labeled probe. Increasing the temperature melts these hybrids. In the next cycle, some of them (approximately 0.1%) hybridize to the appropriate DNA fragment and additional labeled probes are ligated. In this case, differential melting of DNA hybrids with mismatches for both probe sets simultaneously occurs.

In cycle hybridization, before starting cycling,
All components are added at 37 ° C. for T4 and at higher temperatures for thermostable ligase. Then reduce the temperature to 15-37 ° C, incubate the chip for up to 10 minutes,
The temperature is then raised to 37 ° C. or higher for a few minutes and then lowered again. The cycle can be repeated up to 10 times. In one variation, longer ligation reactions (1-3 hours) can be performed using higher optimal temperatures (10-50 ° C) without cycling.

The method described here allows the production of complex chips using standard synthesis and precise oligonucleotide spotting, since relatively few oligonucleotides are required. . For example, all 7-mer oligos (1638
In the case of synthesizing (4 probes), a list of 246 million 14-mers can be measured.

One important variation of the method of the invention is to use two or more differently labeled probes per base array. This is performed for two purposes. That is, multiplexing reduces the number of separately hybridized arrays, or measures a longer list of oligo sequences, such as 3x6 or 3x7. In this case, if two labels are used, the specificity of the three consecutive oligonucleotides can be almost absolute, since the positive site must have sufficient signal for both labels.

A further additional variation is to use a chip containing a BxNy probe (y is 1-4). These chips allow sequencing in different frames. This can also be achieved with an appropriate set of labeled probes, and the F probe and P
Both probes can have several non-specific terminal positions (ie, some elements of terminal degeneracy). Further, a universal base may be used as a part of a linker for binding a probe having a specific sequence to a solid support. This makes the probe more available for hybridization and makes the construct more stable. If the probe has five bases, for example, three universal bases can be used as a linker.

Example 16 Determination of Sequence from Hybridization Data If a given overlapping (N-1) mer is repeated more than once, sequence assembly may be interrupted. Then, two N- nucleotides with different last nucleotides
Any of the mer can be used for sequence extension. This branch point limits the distinct assembly of the sequence.

[0229] Reassembly of the sequence of a known oligonucleotide that hybridizes to the target nucleic acid to produce the complete sequence of the target nucleic acid cannot be achieved in some cases.
This is because some information is lost if the target nucleic acid is not present in a fragment of the appropriate size for the size of the oligonucleotide used for hybridization. The amount of information lost is proportional to the length of the target being sequenced. However, if a sufficiently short target is used, its sequence will be unambiguously determined.

[0230] The expected frequency of overlapping sequences that can interfere with sequence assembly, distributed along a length of DNA, can be calculated. This derivation requires the introduction of definitions of the parameters that must be processed for sequence organization. That is, it is a sequence subfragment (SF). Sequence subfragments result when any part of the sequence of the target nucleic acid starts and ends with an (N-1) mer that is repeated more than once in the target sequence. Thus, a subfragment is a sequence that occurs between two branch points during sequence assembly according to the method of the present invention. The combined length of all subfragments is longer than the actual target nucleic acid due to the overlapping short ends. In general, subfragments cannot be assembled in a linear fashion without additional information. Because they share a (N-1) mer at the end and at the start. A different number of subfragments is obtained for each nucleic acid target depending on the number of repeat (N-1) mers. The number depends on the value of N-1 and the length of the target.

The calculation of probabilities can estimate the interrelation of the two factors. Positive N-
If mer ordering is achieved using overlapping sequences of length N-1 or an average distance of Ao, a fragment of N-1 Lf base lengths is given by the following equation: N _sf = 1 + Ao XKXP (K, L _f ) where K is greater than or equal to 2 and P (K, L _f ) is the number of N-mers present K times on the L _f base fragment Represents the probability. Also, a computer program capable of forming a subfragment from the N-mer content for a given sequence is described in Example 18 below.

The number of subfragments increases with increasing fragment length for probes of a given length. The resulting fragments cannot be uniquely ordered within themselves. Although not complete, this information is very useful for comparative sequence analysis and recognition of functional sequence features. This type of information is called a subarray. Another method for obtaining subsequences is to use only a subset of oligonucleotide probes of a given length.

There is relatively good agreement between the theoretically deduced sequence and the computer simulation of the random DNA sequence. For example, N-1 = 7 [8-mer or 5 '(A, T
, C, G) B ₈ (A, T, C, G) 3 'type of 16 10-mers used], a target nucleic acid of 200 bases would have an average of 3 subfragments . However, due to near-average variability, the library of target nucleic acids should have a 500 bp insert so that less than 1 out of 2000 targets has more than 3 subfragments. Thus, in the ideal case of sequencing long nucleic acids of random sequence, a representative library containing sufficiently short inserts of the target nucleic acid is used. In the case of such inserts, it is possible to reconstruct individual targets by the method of the present invention. Thereafter, the entire sequence of the large nucleic acid is obtained by superimposing the specified individual insert sequences.

To reduce the need for very short fragments (eg, 50 bases for an 8-mer probe), random DNA fragmentation methods such as cloning or overlapping fragments present in random PCR The information contained is used. It is also possible to use pools of short physical nucleic acid fragments. 1 8-mer for sequencing megabase or 11-mer, for example, 5 '(A, T, C , G) N 8 (A, T, C, G) 3' With, 20,000 50 bp fragment Instead of requiring only 2,100 samples are sufficient. This number includes 700 7 kb clones (basic library), 1250 pools of 20 clones of 500 bp (subfragment ordered library) and 150 clones from the jumping (or similar) library. Consists of The algorithm developed (see Example 18) reconstructs the sequence using the hybridization data of these described samples.

Example 17 Algorithm This example is an algorithm for generating a long sequence written in four-letter alphabets from the constituent k-tuple words in a minimum number of randomly determined individual fragments of the starting nucleic acid sequence. Where k is the length of the oligonucleotide probe. This algorithm is primarily intended for use in sequencing by the hybridization (SBH) process. This algorithm is based on the ability to use a pool of physical nucleic acid sequences to define subfragments (SF), informative fragments (IF) and informative fragments.

As noted, subfragments can be caused by branch points in the assembly process that are due to repetition of the target nucleic acid k-1 oligomer sequence. A subfragment is a sequence fragment that occurs between any two repeated words of length K-1 that occurs in the sequence.
Multiple occurrences of the K-1 word cause breaks in the alignment of the K-word overlaps in the alignment process. The interruption leaves the sequence in the form of subfragments. Thus, the distinct segments between the branch points, whose order is uniquely determined, are called sequence sub-fragments.

An informational fragment is defined as a fragment of the sequence determined by the closest end of the overlapping physical sequence fragment.

A certain number of physical fragments can be pooled without losing the possibility of defining information providing fragments. The total length of the randomly pooled fragments is the k used in the sequencing process.
-Depends on tuple length.

This algorithm consists of two main units. The first part is used to generate sub-fragments from the set of k-tuples contained in the sequence. Subfragments can be formed within the coding region of a physical nucleic acid sequence of a certain size, or within an informational fragment defined within a long nucleic acid sequence. Both types of fragments belong to the basic library. This algorithm does not describe the determination of the contents of the k-tuples of the informational fragments of this basic library, ie the stage of production of the informational fragments used in the sequence formation process.

The second part of the algorithm determines the linear order of the sub-fragments obtained for the purpose of reforming the complete sequence of the nucleic acid fragments of the basic library. For this purpose,
A second aligned library of randomly pooled fragments of the starting sequence is used. The algorithm does not include the step of combining the sequences of the base fragments to recreate the entire megabase plus sequence. This may be achieved using the ligation of fragments of the basic library, a requirement for informative fragment formation. Or,
Based on the presence of normal terminal sequences, a search for their overlap may be used, and this algorithm may achieve this after forming the sequence of fragments of the basic library.

[0241] This algorithm uses the predefined k
-It is not necessary to know the number of occurrences of the tuple, nor is it necessary to know that the information of the word of the k-tuple exists at the end of the fragment. The algorithm is implemented with a mixed content of k-tuple words of various lengths. The concept of this algorithm allows operations using k-tuple sets that include false positive and false negative k-tuples. Only in special cases, the content of the pseudo k-tuple mainly affects the integrity and correctness of the resulting sequence. This algorithm may be used to optimize parameters in simulation experiments, as well as in sequence formation in actual SBH experiments, for example in the formation of genomic DNA sequences. In optimizing parameters, the choice of oligonucleotide probes (k-tuples) for practical and convenient fragments and / or the choice of optimal length and number of fragments for defined probes is of particular importance. .

This part of the algorithm plays a central role in the process of forming an array from the contents of the k-tuple. This is based on the inherent alignment of k-tuples with maximum overlap. The major obstacles to sequence formation are specific repeats and false positive and / or false negative k-tuples. The purpose of this part of the algorithm is to obtain the smallest number of longest possible subfragments with the correct sequence. The part of this algorithm consists of one basic step and several control steps. Since some information is only available after the formation of all primary sub-fragments, a two-step process is required.

The main problem with array formation is to obtain repetitive arrays from word contents that, by definition, have no information about the number of occurrences of individual k-tuples. The whole algorithm concept relies on the basics of solving this problem. In principle, there are two opposing approaches: 1) the repetitive sequence is obtained at the beginning of the pSF formation process, or 2) the repetitive sequence is obtained in the subsequent subfragment final alignment process. In the first case, the pSF contains excess sequences; in the second case, they contain sequence deletions in the sequences. The first approach requires the elimination of excess sequences formed, and the second case requires the multiple use of several sub-fragments in the process of final assembly of the sequences.

The difference between the two approaches in the degree of rigor of the inherent overlap law of K-tuples. A less strict rule is that k-tuple X is unambiguous if, and only if, the rightmost k-1 end of k-tuple X is only at the leftmost end of k-tuple Y. That is, it overlaps the k-tuple Y as much as possible. This rule allows the formation of repetitive sequences and the formation of redundant sequences.

The stricter rule used in the second approach has a further caveat: k-tuple X has the rightmost k-1 end of k-tuple X at the leftmost end of k-tuple Y. , And only in that case, and if the leftmost k-1 end of k-tuple Y is not present on the rightmost of any other k-tuple, -Overlap tuples. Algorithms based on stricter rules are simpler and are described herein.

The process of elongation of a given sub-fragment may be such that the k-1 end to the right of the last k-tuple involved is not present on the left of any k-tuple, or that two or more k-tuples are present. Stop if present. If it exists in a k-tuple, this rule of 2
The second part is tested. Furthermore, given a different k-tuple than previously included, the construction of a given subfragment will only terminate at the first leftmost position.
If this additional k-tuple were not present, this condition would fit a unique k-1 overlap and a given sub-fragment would extend one component on the right.

In addition to this basic rule, supplements are used to allow the use of k-tuples of various lengths. The maximum overlap is the shorter k of the overlap pair
-Consists of the length of tuple k-1. PSF formation is performed starting from the first k-tuple obtained from a file in which k-tuples are randomly and independently displayed from their order in the nucleic acid sequence. Thus, the k-tuple of the file is not necessary at the start of the sequence nor at the start of a particular sub-fragment. The process of subfragmentation is performed by aligning k-tuples with unique overlaps defined by the rules described. Each k-tuple used is deleted from the file. At the point when there are no more k-tuples that clearly overlap the last one included, the construction of the subfragment is over and the construction of another pSF begins. Since the formation of the majority of subfragments does not start from their actual onset, the formed pSF is added to a file of k-tuples and is considered a longer k-tuple. Another possibility is to form sub-segments that proceed in both directions from the starting k-tuple. The process ends when no further overlap, ie, extension of any of the sub-fragments, is possible.

The pSF is divided into three groups: 1) the longest and correct sequence subfragment in the case of the correct k-tuple set; 2) the incomplete set and / or false positive k-tuples. Short sub-fragments formed by the use of the largest and clearest overlap rule for sets with some; and 3) pSF with incorrect sequence. The incompleteness of the set of 2) results from false negative results of hybridization experiments, as well as the use of incorrect k-tuple sets. These are formed by false positive and false negative k-tuples, a) misligated subfragments; b) subfragments with incorrect ends; and c) false positives that appear as minimal false subfragments. It may be a k-tuple.

Given the false positive k-tuples, the possibility that there are k-tuples containing more than one wrong base or one wrong base somewhere in the middle, and at the end It may be a k-tuple with the wrong base. Short, erroneous or mis-ligated subfragments are those caused by the latter k-tuple. The former two k-tuples represent an erroneous pSF having a length equal to the length of the k-tuple.

In the case of one false-negative k-tuple, p
SF occurs. If there is one false-positive k-tuple with the wrong base at the left most or right most end, pSF occurs because no clear overlap is possible. If both false-positive and false-negative k-tuples with a common k-1 sequence are present in the file, a pSF results, one of these pSFs containing a false k-tuple at the relevant end.

The process of correcting sequence-corrected subfragments to ligate the clearly ligated pSF occurs after the formation of the subfragments and during the subfragment alignment process. The first step, which consists of excision of misligated pSF and obtaining the final subfragment by unambiguous ligation of pSF, is described below.

There are two approaches for the formation of mislinked subfragments. First, an error occurs when a pseudo k-tuple is found at the construction point of a repeat sequence of length k-1.
Second, the repeats are shorter than k-1. Each of these situations can occur in two variants. In the first variant, one of the repeats represents the end of the fragment. In the second variant, the repeat sequence occurs at any position within the fragment. With regard to the first possibility, the formation of a false bond requires that the file has no k-tuple (false negative). The second possibility requires that both false negative and false positive k-tuples be present in the file. Considering K-1 sequence repeats, it is sufficient to lack one k-tuple if either end has internal repeats. Strictly speaking, the inner repetition needs to lack two. The reason for this is that the ends of the sequence are informatively regarded as endless, linear arrays of false-negative k-tuples. From "smaller than k-1", two or three specific pseudo k-
Only k-2 length repeats requiring tuples would be considered. These are the only cases that will be detected in actual experiments; others are likely to be less frequent.

Recognition of misligated subfragments is more precisely defined when no repetitive sequence is found at the end of the fragment. In this situation, two additional subfragments can be detected, one containing the k-2 sequence at the leftmost end and the other at the rightmost end, which is Also exists. If the repeat sequence is at the end of the fragment, there is only one subfragment containing the k-2 sequence that will erroneously form the subfragment at its leftmost or rightmost end.

The removal of mislinked subfragments by excision is performed according to general rules. That is, if the leftmost or rightmost sequence of the k-2 length of any subfragment is present in any other subfragment, that subfragment is truncated and each of them is a k-2 sequence. Into two sub-fragments. This rule does not cover the rare situation of repeated ends where there is more than one false-negative k-tuple in terms of repeated k-1 sequences. Such mislinked sub-fragments can be identified using overlapping fragments or informational information from the informative fragments of both the basic library and the aligned library. In addition, misligated subfragments remain when more than one false-negative k-tuple occurs at both positions containing the same k-1 sequence. This is a very rare situation because it requires at least four specific pseudo k-tuples. If a given sequence can be obtained by combining sequences shorter than k-2 from the remainder of one subfragment and the beginning of another fragment, these subfragments are excised on the sequence of length k. Additional rules can be introduced to do this.

Due to the strict application of the described rules, some integrity is lost to guarantee the accuracy of the output. Some sub-fragments will be excised without error in ligation as they will match the pattern of mis-ligated sub-fragments. There are several situations with this kind. For example, a fragment may include at least two identical k-1 sequences plus any of the k-1 derived k-2 sequences, or the fragment may have at least two repeated k-2 sequences and a central At least one false negative k-tuple containing the predetermined k-2 sequence, and so on.

The purpose of this part of the algorithm is to reduce the number of pSF to minimize the number of longer sub-fragments with the correct sequence. The formation of unique longer sub-fragments or complete sequences is possible under two circumstances. The first situation concerns a particular order of repeating k-1 words. Some or all of the maximally expanded pSF (the first group of pSF) may have a unique order. For example, the fragment S-R1-a-R2-b-
R1-c-R2-E (where S and E are the start and end of the fragment, a, b and c are different sequences specific to individual subfragments, and R1 and R2 are tandem Repeat
With two k-1 sequences), five subfragments are formed (S-R1, R1-a-R2, R2-B-
R1, R1-c-R2, and RE). These are the original sequences or S-R1-cRb-
R1-aRE can be aligned in two ways. On the other hand, in a fragment having a different order but the same number and type of repetitive sequences, that is, S-R1-a-R1-bRcRE, there is no other sequence including all subfragments. This type of example can only be recognized after the pSF formation process. They demonstrate the need for two steps in the pSF formation process. If the file contains false negative and / or false positive k-tuples, the second situation of the formation of erroneous short subfragments at the position of the non-repetitive k-1 sequence is more important.

For both groups of PSF, the situation consists of two parts. First, false positive k-tuples that appear as the absence of the smallest subfragment are removed. To allow for the formation of the maximum number of linkages, k-tuples of length k with no overlap at either end, longer than ka at one end and longer than kb at the other end All of the subfragments are removed. In our experiments, it was thought that if a and b were 2 and 3, respectively, a sufficient number of false positive k-tuples would be sufficiently eliminated.

[0258] Merging of subfragments that can be uniquely linked is achieved in a second step. The ligation rule states that an overlapping sequence at a suitable end or start of two sub-fragments is not present at the start and / or end of any other sub-fragment, and only then. The two subfragments are clearly linked.

Excludes cases where one subfragment from the pair under consideration has the same start and end point. In this case, concatenation is possible even if another sub-fragment having the same end point exists in the file. The main problem here is the precise definition of the overlapping sequence. If an overlapping sequence that is specific to only one pair of subfragments is shorter than, the same as, or longer than k-2, but additional lengths of overlap sequences of any length longer than k-4 are If present, concatenation is not allowed. Also, the correct ends of pSF and the end after removing one (or several) last base are also considered overlapping sequences.

After this step, some false positive k-tuples (as the smallest subfragment) and some subfragments with incorrect ends remain. Furthermore, in the very rare case where a certain number of somewhat specific pseudo k-tuples are present at the same time, incorrect coupling can occur. These cases are detected and resolved in the subfragment alignment process and in a further control step involving the handling of unremoved "misbound" subfragments.

There are two types of short subfragments obtained. Usually, these sub-fragments can be clearly linked between themselves due to the distribution of the repeating k-1 sequence. This may be done after the pSF formation process and is a good example of the fact that the pSF formation process requires two steps. When using a file containing false positive and / or false negative k-tuples, a short pSF is obtained at the site of the unique k-1 sequence. Considering false-positive k-tuples, k-tuples contain more than one wrong base (or contain one wrong base somewhere in the middle), and a k-tuple at the end. May be included. The formation of short, incorrect (or incorrectly linked) subfragments is caused by the latter k-tuple. A k-tuple of the former kind represents an erroneous pSF having a length equal to the length of the k-tuple.

The purpose of incorporating the pSF portion of the algorithm is to reduce the number of pSF to a minimal number of long subfragments with the correct sequence. For maximum connection,
All sub-fragments of the k-tuple that are longer than ka at one end and longer than the other kb and do not overlap on either end are eliminated. Thus, the majority of false positive k-tuples are rejected. The rule of ligation is that two subfragments can be unambiguously ligated, or that the overlapping termination sequence of the two subfragments is associated with the start and / or termination portion of any other subfragment. It just doesn't exist. Exceptions are subfragments that have identical start and end points. In that case, the concatenation is made provided there is another subfragment with the same end present in the file. The main problem here is the precise definition of the overlapping sequence. With the ligation of false positive and false negative k-tuples, the presence of at least two unique false negative k-tuples at k-1 or k-2 sequence repeats disappears, and some overlap Sequences may be "masked" resulting in error-free and clear concatenation of pSF. To prevent this, the ligation is not performed on terminal sequences shorter than k-2 and in the presence of special overlapping sequences longer than k-4. The overlap array is
It is revealed from the end of pSF, ie one or only a few final bases have been removed.

In very rare situations, the presence of a certain number of certain false-positive and false-negative k-tuples will leave some sub-fragments with erroneous ends and some false-positive k-tuples.
-Tuples (as minimal sub-fragments) may remain, ie misligations may occur. These problems are identified and resolved in the subfragment alignment process and in a further control step, along with the management of uncut, mislinked subfragments.

The subfragment alignment process is similar to their formation process. If one considers the subfragments as long k-tuples, the alignment is achieved by their distinct connection by terminal overlap. The information platform for unambiguous ligation is the division of the subfragments that occur in the fragments of the basic library into groups corresponding to the segments of those fragments. The method is similar to a biochemical solution to this problem based on hybridization with a long oligonucleotide having an associated linking sequence.
The linking sequence is formed as a sub-fragment using the appropriate k-tuple set segment of the basic library fragment. Relevant segments are defined by aligned library fragments that overlap each fragment of the base library. The shortest segment is the informative fragment of the aligned library. The longest is the entire overlapping portion of several adjacent informative fragments or fragments corresponding to the alignment and base library. To reduce the number of samples to be separated, the fragments of the aligned library are randomly pooled and the unique k-tuple content is determined.

By using multiple fragments of the aligned library, very short segments are formed, thus reducing the chance of many occurrences of the k-1 sequence responsible for the formation of subfragments. Furthermore, the longest segments consist of various regions of a given fragment of the basic library and do not contain any repeating k-1 sequences. In every segment, a joining sequence (joining sub-fragment) is formed into certain sub-fragment pairs of a given fragment. The alignment process consists of three steps: (1) formation of a k-tuple contained in each segment; (2) formation of a sub-fragment of each segment; and (3) ligation of the segment with a sub-fragment. The first segment is defined as the significant intersection and difference between the k-tuple content of a given fragment of the base library and the k-tuple content of the aligned library pool. The second (short) segment is defined as the common and different parts of the k-tuple contents of the first segment.

[0266] Both the intersection and the difference have the problem of aggregating both false positive and false negative k-tuples. False-positive k-tuples occur randomly in both sequences that are not the relevant overlap region, and false-negative k-tuples in the starting sequence accumulate in the intersection (overlap). On the other hand, most false-positive k-tuples of any of the starting sequences are not covered by the intersection. This is an example of using information from overlapping fragments to reduce experimental errors from individual fragments. False k-tuples accumulate in the differences for another reason. The set of false negatives in the original sequence is included in the false positives in the intersection and false positives for those k-tuples that are not accidentally included in the intersection, i.e., the intersection is false negative in the intersection. Extended to a set of If the starting sequence contains 10% false negative data, the first and second intersections will contain 19% and 28% false negative k-tuples, respectively. On the other hand, if the base fragment and pool had a length of 500 bp and 10,000 bp, respectively, a false positive with an expected value of 77 would be expected. However,"
It is possible to regenerate most of the "lost" k-tuples and to eliminate most of the false positive k-tuples.

First, it is necessary to determine the basic content of a k-tuple of a predetermined segment as a common part of the content of a predetermined pair of k-tuples. This is followed by the start k-tuple contents of the common part containing k-1 at one end and the k- + sequences occurring at the ends of the two k-tuples of the basic set at the other end. -Check that it contains tuples. This is done before the process creates differences that prevent the accumulation of false positives.
Then we apply the same type of k-tuple extension set to the different parts with the difference that the borrowed one is from the common part. All k borrowed
-Tuples are excluded from the common file as false positives.

The intersection, ie, the set of common k-tuples, consists of each pair (basic fragment) X (
Pool of aligned libraries). If the number of k-tuples in the set is significant, it is increased by false negatives according to the rules described. The first set of differences is obtained by subtracting the common set obtained from the given base fragment set. False-negative k-tuples are borrowed from the common set according to the rules described, applied to the difference set, and simultaneously removed from the common set as false-positive k-tuples. If the base fragment is longer than the pooled fragment, this difference would mean two different segments that would somewhat reduce its utility in further steps. The first segment is all that yields the intersection and divergence of pairs (basic fragments) X (pool of ordered library) containing a significant number of k-tuples. No.
The k-tuple set of the second segment is obtained by comparing all possible pairs of k-tuple sets of the first segment. The two differences are defined from each pair resulting in an intersection with a significant number of k-tuples. Most of the information available from the overlapped fragments is recovered in this step so that little is gained from the third round forming the intersection and dissimilarity.

(2) The formation of the sub-fragments of the segments is performed in the same manner as described for the fragments of the basic library.

(3) The method of ligation with sub-fragments determines successively the sub-fragment pairs that are correctly ligated among the sub-fragments of a given basic library fragment having several overlapping ends. Consisting of things. In the case of four related fragments, two of which contain the same start and two have the same termination, there are four different pairs of ligable subfragments. In general, two are correct and two are incorrect. To find the correct one, the presence of each pair of joining sequences is tested on the sub-fragments resulting from all the first and second segments of a given basic fragment. The length and position of the linking sequence are chosen to avoid accidental sequence interference. They are either k + 2 or longer and additionally contain at least one element 2 by the overlapping sequence of both subfragments of a given pair. Ligation is only made if two ligation sequences are found and the other two are not present. The two concatenated sub-fragments return the previous sub-fragment in the file, and the process repeats periodically.

[0271] Repetitive sequences are formed in this step. This means that several sub-fragments are included in the sub-fragments that are ligated more than once. They will be recognized by finding related ligation sequences that link one sub-fragment with respect to two different sub-fragments.

The mislinked recognition resulting from the process of constructing pSF and incorporating pSF depends on whether the sequence of a subfragment of a given base library is present in the sequence of the subfragment formed by a segment of that fragment. Based on the test. Sequences at misligated positions will not be known to indicate misligated subfragments.

In order to form a more complete sequence without errors, in addition to the described three steps of aligning the subfragments, some additional control steps or steps applicable to a particular sequence may be required.

By comparing the contents of the k-tuples in segments and sub-fragments, a determination is made as to which sub-fragment belongs to which segment. Due to errors in the k-tuple contents (due to the first error in the pool and statistical errors due to frequent k-tuple occurrences), exact segmentation of sub-fragments is not possible. Thus, instead of an "all or nothing" split, the opportunity arising from a given segment for each sub-fragment is determined. This probability is a function of the length of the k-tuple, the length of the sub-fragments, the length of the fragments of the aligned library, the size of the pool, and the percentage of false k-tuples in the file: P (sf, s ) = (Ck-F) / Lsf, where Lsf is the length of the sub-fragment, Ck is the number of common k-tuples for a given sub-fragment / segment pair, and F is the length of the k-tuple. Parameters including the relationship between the base library fragments, pool size, and% error).

[0275] Sub-fragments of a particular segment are treated as redundant short pSFs and undergo a well-defined concatenation process. Because of the overlapping ends, a clear ligation decision is in this case
Slightly different. Note that the precision of explicit concatenation is controlled according to these sub-fragment concatenations of other segments. After ligation of the different segments, all the resulting sub-fragments are incorporated together, the shorter ones of the longer sub-fragments are eliminated, and the remaining ones undergo the usual ligation process. If, after explicit ligation, the sequence is not completely reconstructed, the sub-fragment segmentation and ligation process is repeated with the same or less stringent criteria for the probability of belonging to a particular segment.

Some information is not used when using strict criteria to define a clear overlap. Instead of the complete sequence, several sub-fragments are obtained that define many possibilities for a given fragment. When using less stringent criteria,
An accurate and complete sequence is provided. Certain situations, such as incorrect ligation, can result in complete but incorrect sequences, or "large" subfragments with no linkage between them. Thus, each fragment of the basic library yields a) several possible solutions in terms of accuracy and b) the most promising accurate solution. Also, in very rare cases, errors during the subfragmentation process, or a fixed percentage of likelihood to produce, yield a definite solution, or one, the most promising solution. In these cases, the sequences are incomplete, or a clear solution can be obtained by comparing these overlapping data with other overlapped fragments of the basic library.

The described algorithm was tested on a randomly generated 50 kb sequence containing 40% GC simulating the GC content of the human genome. In the middle of this sequence, all various and some other repetitive sequences with a total length of about 4 kb were inserted. In order to simulate the in vitro SBH test, the following operations were performed to prepare appropriate data.

The position of -60 5kb overlapping "clones" was randomly defined to simulate the preparation of the basic library:-The positions of 1000 500bp "clones" were randomly defined and aligned live The production of the rally was simulated. These fragments were obtained from the sequence. A random pool of 20 fragments was generated: the pool k-tuple set was determined and placed on the hard disk. These data were used in the subfragment alignment step: instead of clones of the same density, 4 million clones of the base library and 3 million clones of the aligned library are used for the whole human genome. A total of 7 million clones are several times lower than clones several kb long for random cloning of almost all genomic DNA and sequencing by gel-based methods.

From the data at the beginning and end of the 5 kb fragment, 117 “informational fragments”
Was determined to be in the sequence. After this, the set of overlapping k-tuples that make up the signal "informational fragment" was determined. Only a subset of the k-tuples in the predetermined list were used. The list includes 8-mer 65%, 9-mer 30%
And 5% from 10 to 12-mer. Based on these data, a subfragment formation and alignment process was performed.

Testing of the algorithm was based on data from two simulated tests. 50 pieces of information, including 26 informational fragments (approximately 10,000 bp), including 100% accurate datasets (> 20,000bp) and 10% pseudo k-tuples (5% positive and 5% negative) The sequence of the donated fragment was regenerated.

In the first test, all sub-fragments were correct and 1 in 50 informative fragments
Only one sequence was not completely regenerated and remained as five subfragments. Analysis of the position of the overlapping fragments in the alignment library indicated that they lacked information on the specific alignment of the five subfragments. The subfragments can be ligated in two ways based on the overlapping ends, 1-2-3-4-5 and 1-4-3-2-5. The only difference is that the positions of subfragments 2 and 4 are interchanged. Because subfragments 2, 3, and 4 are relatively short (all about 100 bp), there are relatively many opportunities in this case, ie, when there are no fragments of the aligned library starting or ending at the subfragment 3 region. Existed and happened.

In order to simulate actual sequencing, many tests included some erroneous (“hybridization”) data as input. In oligomeric hybridization tests, the only event that produces unreliable data under the conditions of interest is mismatched versus perfectly matched hybridization. Therefore, in simulations, only those k-tuples in which one element at either end was different from the real one were considered false positives. These “wrong” sets are made as follows. A subset of 5% false positive k-tuples is added to the original set of k-tuples of the informative fragment. False positive k-tuples are created by randomly selecting a k-tuple from the set, copying it, and then partially changing the nucleotides at its start or end. After this, we remove a randomly selected subset of the 5% k-tuple. Thus, the most complicated situation in which the correct k-tuple is replaced by a k-tuple having the wrong base at the end occurs with a statistically guessed number.

The formation of the described k-tuple set results in 10% incorrect data. This value changes in every case because the k-tuple to be copied, changed, and deleted is randomly selected. Nevertheless, this percentage exceeds the unreliable data volume in actual hybridization tests by three to four times. The 10% error derived will result in a two-fold increase in the number of sub-fragments in both the basic library fragment (the basic library information providing fragment) and the segment. Approximately 10% of the final subfragment has the wrong base at the end, as expected in a k-tuple set including false positives (see formation of the first subfragment). There were no cases where the sub-fragments were misligated or the sub-fragments had the wrong sequence. Four of the 26 informative fragments tested in the alignment process did not reproduce the complete sequence. In all four cases, the sequences were obtained as several long and several short subfragments contained in the same sequence. This result indicates that the algorithm principle is driving a high percentage of incorrect data.

[0284] Successful sequence formation from k-tuple inclusions is described in terms of completeness and accuracy. In the formation process, two special situations are: 1) some information pieces are dropped in the resulting sequence, but it is known where the ambiguities are and what types they belong to, and 2) It can be clarified that the regenerated sequence obtained does not match the sequence from which the k-tuple content is generated, and that the error cannot be detected. Assuming that the algorithm expands to the theoretical limit, only the first simulation can be performed to use the correct k-tuple. At that point, the imperfection causes a certain number of subfragments to become unclearly aligned, and furthermore to the problem of determining the exact length of a single sequence, ie, the number of complete tandem repeats.

It is believed that the false k-tuple results in an incorrect sequence. The error is not due to the lack of algorithm, but to the fact that certain k-tuple inclusions clearly represent sequences that differ from the first. Based on the type of spurious k-tuple present in the file, three errors can be defined. False negative k-tuples (without false positives) cause "deletions". False positive k-tuples cause "extensions (unequal crossings)". False positives with false negatives are responsible for "insertions", alone or in combination with "deletions". Deletions are caused when all k-tuples (or most of them) between the two possible starting portions of a subfragment are false negative. Since every position in the sequence is defined by a k-tuple, in the general case k consecutive false negatives are required if a deletion occurs. (This situation occurs after each 108 elements, with 10% false negatives and k = 8.) This situation is very rare, even in mammalian genome sequencing using a random library containing 10 genomic equivalents. .

A sequence extension caused by a false-positive k-tuple is a special case of “insertion” because the ends of the sequence can be considered a circulating linear array of false-negative k-tuples. It is considered a group of false positive k-tuples that yield subfragments longer than one k-tuple. Such a situation would be detected if subfragments were formed in overlapping fragments, such as random physical fragments of an ordered library. Insertions in place of insertions or deletions can occur as a result of the special combination of false positive and false negative k-tuples. In the first case, the number of consecutive false negative k-tuples is lower than k. In both cases, several overlapping false positive k-tuples are required. Since the requirements on the number and specificity of the pseudo k-tuple are so simple, insertions and deletions are largely theoretically feasible, without any real significant impact.

In all other situations where the number of false positives and / or false negatives does not meet the theoretical requirement of a minimum number, the integrity of the erroneous sequence of k-tuple inclusions is reduced. It seems to lower.

SBH, nucleic acid samples are sequenced by exposing the sample to a support-bound probe of known sequence, and a labeled or lysed probe. If probe ligase is introduced into the probe and sample mixture, i.e., if the support has bound and labeled probes hybridized back to back along the sample, the two probes are chemically bound by the action of ligase. Like. After washing, only chemically bound support binding and labeled probes are detected by the presence of labeled probes. Knowing the identity of the support-bound probe and the identity of the labeled probe at a particular location in the array, a portion of the sample sequence can be determined by the presence of the label at the point of the array in a format with three substrate samples. . Not by chance and not by chance, but not all sequences of the bound probe pair will overlap maximally and the sequence of the sample will be reconstructed. The unsequenced sample will be a nucleic acid fragment or a 10 base pair ("bp") oligonucleotide. The sample is preferably between 4 and 1000 bases in length.

The length of the probe is a fragment with a length of less than 10 bases, preferably between 4 and 9 bases. Thus, an array of support-bound probes will have all oligonucleotides of a given length, or only those oligonucleotides selected for a particular test. When using predetermined lengths of all the oligonucleotides, the center of the oligonucleotide number 4 ^N (N is the length of the probe) will be calculated by.

Example 18 Use of a Sequencing Chip If ligation is employed in the sequencing method, ordinary oligonucleotide chips cannot be immediately reused. We believe that this problem can be overcome in various ways.

Ribonucleotides can be used for the second probe, Probe P, so that the probe can be removed in a subsequent RNAse treatment. In RNAse treatment, an endoribonuclease that specifically attacks single-stranded RNA 3 'to the pyrimidine residue and cleaves the phosphate bond to adjacent nucleotides
Use A. The final product is an oligonucleotide with pyrimidine triphosphate and terminal pyrimidine triphosphate. RNAse A works in the absence of cofactors and divalent cations.

To use RNAse, the chip is generally incubated in a suitable RNAse-containing buffer as described in Sambrook et al. (1989; incorporated herein by reference). 37-50 μl RNAse-containing buffer per 8x8 mm or 9x9 mm array
It is preferred to use at 60C for 10-60 minutes. Thereafter, it is washed with a hybridization buffer.

Although not broadly applicable, uracil bases can be used in certain embodiments, as described in Craig et al. (1989; incorporated herein by reference). Disruption of the ligated probe combination (resulting in a reusable chip) is achieved by digestion with the E. coli repair enzyme uracil-DNA glycosylase (removing uracil from DNA).

[0294] Alternatively, a bond that can be specifically cleaved can be formed between the probes, and the bond can be cleaved after detection. For example, this is described in Shabarova et al. (1991) and Dolinnaya
(1988) (both references are hereby incorporated by reference).

[0295] Shabarova et al. (1991) describe condensing oligodeoxyribonucleotides using cyanogen bromide as a condensing agent. In their one-step chemical ligation reaction, the oligonucleotide is heated to 97 ° C., gradually cooled to 0 ° C., and then 1 μl of 10 mM BrCN in acetonitrile is added.

[0296] Dolinnaya et al. (1988) describe how to incorporate phosphoramidites and pyrophosphate internucleotide linkages into DNA duplexes. They have also used chemical ligation methods to modify the sugar phosphate backbone of DNA using water-soluble carbodiimide (CDI) as the coupling agent. Selective cleavage of the phosphoramidite bond requires contact with 15% CH ₃ COOH at 95 ° C. for 5 minutes. For selective cleavage of the pyrophosphate bond, a pyridine-water mixture (9: 1) and freshly distilled (
Contact with CF ₃ CO) ₂ O is required.

Example 19 Diagnosis of Known Mutations-Scoring or Resequencing of the Complete Gene In simple cases, the goal was to find out if a selected known mutation occurred in a DNA segment. It may be. Less than 12 probes are sufficient for this purpose, for example, 5 probes positive for one allele, 5 positive for the other, and 2 negative for both. Because the number of probes scored for each sample is small, it is possible to analyze many samples in parallel. For example, using 12 probes in three hybridization cycles, 96 different genomic loci or gene segments from 64 patients can be analyzed on a single 6x9 inch membrane. The membrane contains a 12x24 sub-array, each sub-array having 64 dots representing the same DNA segment from 64 patients. In this example, a sample may be prepared in 64 96-well plates. There can be one patient per plate and one DNA segment to be analyzed per well. Samples from 64 plates may be spotted every quarter of the same membrane to make four replicas.

A set of 12 probes is selected by single channel pipetting or a single pin transfer device (or an array of individually controlled pipettes or pins) for each of the 96 segments. The selected probes can be arranged in twelve 96-well plates. If the probe is not pre-labeled, it may be labeled, and the probes from the four plates may then be mixed with the hybridization buffer and preferentially added to the subarray by a 96-channel pipetting device. After one cycle of hybridization, the previously applied probe can be removed by incubating the membrane at 37 ° C to 55 ° C, preferably in undiluted hybridization or wash buffer. It is.

The probability that a probe positive for one allele is positive and a probe positive for the other allele is negative can be used to determine which of the two alleles is present. it can. This duplication scoring scheme allows for a certain level of error (about 10%) in the hybridization of each probe.

In particular, where less redundancy is sufficient (eg, if one or two probes demonstrate whether one of the two alleles is included in the sample)
Many alleles can also be scored using an incomplete set of probes. For example, using a set of 4000 8-mers, there is a 91% chance of finding at least one probe that is positive for one of the two alleles for a randomly selected locus. . An incomplete set of probes may be optimized to reflect the G + C content and other bias of the test sample.

When sequencing a complete gene, the gene may be amplified with an appropriate number of segments. A probe set (about one probe per 2 to 4 bases) may be selected and hybridized for each segment. These probes can determine where in the test segment the mutation is. Segments in which one or more mutation sites are detected (ie, subarrays containing these segments) can hybridize to additional probes and find the correct sequence at the mutation site. When the DNA sample was tested with each of the second 6-mers, the mutation was localized at a position surrounded by the positively hybridized probes TGCAAA and TATTCC and covered by the three negative probes CAAAAC, AAACTA and ACTATT. If so, the mutated nucleotide should be A and / or C present at that position in the normal sequence. This change may be caused by a single base mutation or by the deletion or insertion of one or two nucleotides between bases AA, AC or CT.

One approach is to select a probe in which the positively hybridized probe TGCAAA has been extended to the right by one nucleotide and a probe in which the positively hybridized probe TATTCC has also been extended to the left by one nucleotide. It is to do. These eight probes (GCAAAA, GCAAAT, GCAAAC, GCAAAG and ATATTC
, TTATTC, CTATTC, GTATTC) are used to determine the two nucleotides in question.

The most likely guess for the mutation can be determined. For example, A is G
It turns out that it mutates to. There are two interpretations that satisfy these results. one,
The only change is to replace A with G. In addition to this change, a certain number of bases are inserted between newly determined G and subsequent C. That is. If the result with the cross-linking probe is negative, first by at least one cross-linking probe containing the mutation position (AAGCTA), then by an additional 8 probes CAAAGA, CAAAGT, CAAAGC, CAAAGG and ACTATT, TCTATT, CCTATT , GCTATT may be used to confirm these options. To select a mutation elucidation probe,
There are many other ways.

In the case of diploid, heterozygotes can be identified by comparing the scores individually for the test sample and the homozygous control (see above). If the segment covered by a contiguous probe is mutated on one of the two chromosomes, then some of these contiguous probes are expected to have about one-half the signal.

Example 20 Identification of Genes (Mutations) That Cause Genetic Disorders and Other Traits A universal set of long probes (8-mer or 9-mer) is used on an immobilized array of samples. Thus, a DNA fragment of about 5 to 20 kb can be sequenced without subcloning. In addition, the speed of sequencing can easily be about 10,000,000 bp / day / hybridizer. This capability allows iterative resequencing of a large fraction of human genes or human genomes from individuals of scientific or medical interest. To resequence 50% of human genes,
100,000,000 bp base pairs are checked, but can be performed at a reasonable cost in a relatively short period of time.

This superior resequencing ability can be used in several ways to identify genes encoding mutations and / or disorders or any other trait. Basically, mRNA from a specific tissue or genomic DNA of a patient with a specific disorder (
which can be converted to cDNA) may be used as a starting material. Genes or genomic fragments of appropriate length can be prepared from both DNA sources by cloning or in vitro amplification techniques (eg, PCR). When using cloning, a minimal set of clones to be analyzed may be selected from the library and then cloned. This operation is efficient when a small number of clones longer than 5 kb are selected, since only a small number of probes need to be hybridized. Cloning can approximately double the amount of hybridization data, but does not require tens of thousands of PCR primers.

In one refined procedure, DNA is cut like GACGC (N5 ′) / CTGCG (N10 ′)
Gene fragments or genomic fragments can be prepared by restriction cleavage with an enzyme such as HgaI. The overhang consisting of 5 bases differs for each fragment. One enzyme produces a fixed fragment for a certain number of genes. By cleaving cDNA or genomic DNA with a plurality of enzymes in separate reactions, each target gene can be appropriately cut out. In one approach, the cut DNA is fractionated by size. The DNA fragment prepared in this manner (and additionally treated with exonuclease III, which removes nucleotides one by one from the 3 'end and increases the length and specificity of the end), is added to a tube or multiwell. It may be distributed on a plate. From a relatively small set of DNA adapters with common portions and variable length overhangs, adapter pairs can be selected for each gene fragment that needs to be amplified.
These adapters are ligated and then PCR is performed with universal primers. 100
It is possible to generate 1,000,000 pairs from zero adapters, and thus, it is possible to specifically amplify 1,000,000 different fragments under the same conditions with a universal primer pair complementary to the common end of the adapter. is there.

If the DNA difference is found repeatedly in multiple patients, such sequence changes may be nonsense or may alter the function of the corresponding protein, and therefore, the mutated gene Can cause disability. By analyzing a significant number of individuals with a particular trait, functional allelic variants of a particular gene can be associated with a particular trait.

Using this approach, there is no need for expensive genetic mapping for a wide range of families, and certain utilities are found without such genetic data or genetic material.

Example 21 Scoring of Single Nucleotide Polymorphism in Gene Mapping The technique disclosed in this application is suitable for efficient identification of genomic fragments having single nucleotide polymorphism (SNUP). Cloning or in 10 individuals
By applying the described sequencing method to a large number of genomic fragments of a known sequence that can be amplified by in vitro amplification, a sufficient number of DNA segments with SNUP can be identified. Fragments with polymorphism are further used as SNUP markers. These markers may be already mapped (for example, corresponding to the mapped STS), or may be mapped by the screening method described below.

[0311] SNUP can be scored by amplifying markers and arranging them in an array of subarrays for each individual belonging to the relevant family or population. The subarray contains the same markers amplified from the subject. For each marker, as in the diagnosis of known mutations,
A set consisting of no more than six probes and no more than six probes positive for the other allele can be selected and scored. By significantly associating a single marker or group of markers with a disorder, the chromosomal location of the causative gene can be determined. Due to high throughput and low cost, thousands of markers can be scored for thousands of individuals. Because of this amount of data, the location of genes can be determined at a resolution level of less than 1,000,000 bp, and the locations of genes involved in polygenic diseases can be determined. By sequencing specific regions from affected normal and diseased individuals and scoring for mutations, localized genes can be identified.

For amplification of a marker derived from genomic DNA, PCR is preferred. Each marker requires a specific primer pair. Existing markers may be diverted or new markers may be defined that can be prepared by cutting genomic DNA with an HgaI type restriction enzyme and ligating with an adapter pair.

Amplifying the SNUP marker or spotting as a pool can reduce the number of independent amplification reactions. In this case, more probes are scored per sample. If 4 markers are pooled and spotted on 12 replica membranes, 48 probes (12 per marker) can be scored in 4 cycles.

Example 22 Detection and Verification of Identity of DNA Fragments Restriction cleavage, cloning or multiple amplifications in vitro (eg, P
CR) can be identified experimentally. Identification can be performed by verifying whether a DNA band of a specific size is present as a result of gel electrophoresis analysis. Alternatively, specific oligonucleotides can be made and used to verify the DNA sample in question by hybridization. The approach developed herein allows more efficient identification of a large number of samples without having to create a specific oligonucleotide for each fragment. The set of positive and negative probes can be selected from a universal set of each fragment based on the known sequence. Probes that are selected to be positive can usually form one or several overlapping groups, while negative probes diffuse to all insertion sites.

The techniques of the present invention can be used to identify STSs during the process of mapping YAC clones. Each of the STSs can be tested on about 100 YAC clones or a pool of YAC clones. The DNA obtained from these 100 reactions is spotted on one subarray. The different STSs may be sequential sub-arrays. In several cycles of hybridization, D
For each of the NA samples, a discriminating characteristic can be formed that proves with certainty or does not prove whether a particular STS is present in a given YAC clone.

In order to reduce the number of individual PCR reactions or the number of individual spot samples, respectively, several STSs may be amplified simultaneously in one reaction, or
Samples may be mixed. In this case, more probes must be scored per dot. STSs storage is independent of YACs being stored and can be used for one YACs or a pool of YACs. When several probes labeled with different colors are hybridized together,
This proposal is particularly attractive.

In addition to determining whether a DNA fragment is present in a sample, the intensity of hybridization of several separate probes or one or more pools of probes can be used to estimate the amount of DNA. By comparing the intensity obtained with that of a control sample having a known amount of DNA, the amount of DNA in all spotted samples is determined simultaneously. Since only a few probes are required to identify a DNA fragment, and N probes can be used for DNA having a base length of N bases, it is sufficient for the use of the present invention to identify any DNA segment. A simple probe does not require many sets. From the 1000 8-mers, an average of about 30 full-length matching probes can be selected for a 1000 bp fragment.

Example 23 Identification of Infectious Pathogens and Variants Thereof DNA-based tests for detecting viruses, bacteria, fungi and other parasites in patients are usually more reliable than other methods. High and cheap. A major advantage of DNA testing is the ability to identify specific strains and variants, and ultimately to apply more effective treatments.

The presence of 12 known antibiotic resistance genes in bacterial infections can be tested by amplifying these genes. 2 amplification products from 128 patients
Spotted on one subarray, then 24 subarrays of 12 genes were
It can be repeated four times with a x12 cm membrane. For each gene, 12 probes can be selected for positive and negative scoring. Hybridization can be performed under three cycles. In these tests, a fairly small set of probes is most likely universal. For example, out of 1000 8-mers, an average of 30 probes are positive for the 1000 bp fragment and 1
Zero positive probes are usually sufficient for reliable identification. As described in Example 9, several genes can be amplified and / or spotted together to determine a given amount of DNA. The amount of the amplified gene can be used as an indicator of the level of infection.

Another example concerns the possibility of sequencing one gene or the whole genome of the HIV virus. Due to the speed of diversification, this virus presents great difficulties in selecting the optimal treatment. DNA fragments from viruses isolated from up to 64 patients can be amplified and resequenced by the described techniques. Based on the sequence obtained, the optimal treatment can be selected. If there is a mixture of two viral species, one of which has the basic sequence (as in the case of the heterozygous type), the hybridization score is determined by comparing the hybridization score with the other samples, especially the control containing only the basic viral species. It can be identified by quantitative comparison with the sample score. About 3 times the score is obtained with 3 to 4 probes covering one mutation site of the two virus species present in the sample (see above).

Example 24 Forensic identification and parent-child identification Polymorphisms in the sequence make each genomic DNA unique. This allows analysis of blood or other bodily fluids or tissues taken from crime scenes and comparison with samples of suspected criminal. A sufficient number of polymorphic sites are scored to form a unique signature of the sample. SBH readily scores single nucleotide polymorphisms to produce such discriminating properties.

One set of DNA fragments (10-1000) from a sample and a suspect can be amplified. Samples representing one fragment and DNAs from suspects are spotted on one or several subarrays, and each subarray can be replicated four times. In three cycles, twelve probes can determine the presence of allele A or B in each of the samples containing suspects for each DNA locus. By matching the sample and the pattern of the suspect, the suspect of the crime can be found.

The same approach can be applied to proof of the identity of a child's parent. Prepare DNA,
Allelic A or B patterns can be determined by amplifying polymorphic loci from children and adults; hybridization for each. The patterns obtained and the comparison of the positive and negative controls will help determine family relationships. In this case, identification requires that only the significant portion of the allele be paired with one parent. A high number of scoring loci allows avoiding the statistical error of this approach or the shielding effect of de novo mutations.

EXAMPLE 25 Assessing Genetic Diversity of Populations or Species and Biodiversity of Ecological Status A significant number of loci (eg, some genes or mitochondrial DN
By measuring the frequency of allelic variation for A), various types of results can be obtained, such as those relating to the genotype, history and evolution of one population or the effects of the environment on disease or potential for extinction. Resequencing the full length of several loci by testing for certain known alleles or to form small mutations or de novo mutations that can reveal the presence of mutagens in the environment By doing so, these evaluations can be performed.

Alternatively, biodiversity in the microbial world can be investigated by resequencing an evolutionarily conserved DNA sequence, such as the gene for ribosomal RNAs or the gene for highly conserved proteins. . DNA can be prepared from the environment and specific genes can be amplified using primers corresponding to the conserved sequences. The DNA fragment is preferably cloned into a plasmid vector (or diluted to a concentration of one molecule per well of a multiwell plate and then amplified in vitro). Clones generated in this manner can be resequenced as described above. Two types of information are obtained. As a first step, a list of different species can be formed, while the densities of the various individuals can be defined. Other information can be used to determine the impact of ecological factors or pollution on ecosystems. It becomes clear whether some species will be eradicated or if the abundance ratio of species will be altered by contamination. The method of the present invention is also applicable for sequencing fossil DNAs.

Example 26 Detection or Quantification of Nucleic Acid Species DNA or RNA species can be detected and quantified using an unlabeled probe immobilized on a substrate or a probe pair containing a labeled probe in solution. Species can be detected and quantified by exposure to unlabeled probe in the presence of labeled probe and ligase. Specifically, the formation of an extended probe on the nucleic acid skeleton of the sample by the ligation of the labeled and unlabeled probes indicates that the species to be detected is present. Thus, after removal of the unligated labeled probe, the presence of the label at a particular location on the array on the substrate indicates the presence of the species in the sample, while the quantification of the label indicates the expression level of the species.

Alternatively, one or more unlabeled probes can be arranged on a substrate as a first member of a pair with one or more labeled probes to be introduced in solution. According to one method, multiplexing of the labels on the array can be performed using a dye that fluoresces at an identifiable wavelength. In this method, the mixture of the cDNA applied to the array and the pair of labeled and unlabeled probes specific for the species to be identified is c
The presence and expression level of the DNA species can be investigated. According to a preferred embodiment, the method of the present invention is performed to select cDNA and unlabeled probe pairs that contain sequences that overlap the sequence of the cDNA to be detected.
Can be arranged.

By selecting a probe and including a selected probe pair present in the composition in combination only with the target pathogen genome in the composition, the presence of the particular pathogen genome can be detected and quantified. it can. Thus, one probe pair need not be specific to the genome of a pathogenic organism, but the combination of pairs is specific.
Similarly, when detecting or sequencing cDNAs, certain probes may be identified as cDNs.
It may happen that it is not specific to As or other species. Nevertheless, the presence and quantification of a particular species is determined by the result that placement of the selected probe combination at a separate array location indicates the presence of the particular species.

Infectious bacteria having about 10 kb or more of DNA can be analyzed by polymerase chain reaction (PCR).
Alternatively, detection can be performed using a support-bound detection chip without using other target amplification techniques. According to another method, the genome of infectious bacteria, including bacteria and viruses, is assayed by amplifying one target nucleotide sequence by PCR and detecting the presence of the target by hybridization of a labeled probe specific for the target sequence. You. Because such assays are specific to only one target sequence, it is necessary to amplify the gene by a method such as PCR to provide sufficient target for a detectable signal.

According to this example, a solid phase detection chip comprising an array of multiple different immobilized oligonucleotide probes specific to the infectious agent of interest is prepared, characterized by infectious agents by format 3-type reaction. An improved method for detecting a specific nucleotide sequence is provided. One dot containing a mixture of many unlabeled probes complementary to the target nucleic acid enriches the species-specific label at one location, thereby improving sensitivity over diffusion or single probe labeling methods. You. Such multiple probes may be overlapping sequences of the target nucleotide sequence,
Non-overlapping sequences as well as non-adjacent sequences may be used. Such probes preferably have a length of about 5 to 12 nucleotides.

The nucleic acid sample exposed to the probe array and the target sequence present in the sample hybridizes with the multiple immobilized probes. The pool of multi-labeled probes selected to specifically bind to the target sequence adjacent to the immobilized probe is then applied along with the sample to an array of unlabeled oligonucleotide probe mixtures. Then
A ligase enzyme is applied to the chip to ligate adjacent probes to the sample. Next, the detection chip is washed to remove the non-hybridized and non-ligated probe and sample nucleic acid, and the presence of the sample nucleic acid can be determined based on the presence or absence of the label. This method provides a reliable sample detection method in which the molar concentration of the sample drug is reduced by about 1000 times.

In yet another embodiment of the present invention, which comprises a chromogenic multilabel, an enzymatic multilabel or a radioactive multilabel per se, or as such is susceptible to specific binding by another multilabeled probe agent. The labeled probe signal can be amplified, such as by providing a common end with the free probe. In this method, the second signal amplification can be performed. Labeled or unlabeled probes can be used for a second round of amplification. In the second amplification, if the DNA sample having a multiplex label is long, the amplification intensity signal increases by 10 to 100 times, and the total amplification signal increases by 100,000 times. By using both aspects of this example, about 10
A positive result of probe-DNA ligation can be obtained with a 0000-fold intensity signal.

According to yet another aspect of the invention, an array or superarray consisting of a complete set of probes, for example, 4096 6-mer probes, can be generated. Arrays of this type are universal in the sense that any nucleic acid species can be identified, or partially sequenced, of any nucleic acid species. Each spot in the array contains one probe species or a mixture of probes, eg, a mixture of N (1-3) B (4-6) N (1-3) species synthesized in one reaction. (N indicates all four nucleotides, B indicates one specific nucleotide, and the associated numbers are in the range of the number of bases, ie, 1-3 means "1-3 bases" These mixtures provide a stronger signal of nucleic acid species present at low concentrations by collecting signals from different parts of the same long nucleic acid species molecule. The universal set of probes is subdivided into a number of subsets and spotted with the sample and labeled probes as a unit array separated by a barrier that prevents diffusion of the hybridization buffer.

To identify nucleic acid species having a known sequence, one or more oligonucleotide sequences can be selected that include both an unlabeled immobilized probe and a labeled probe solution. The labeled probes are synthesized or selected from a complete set of pre-synthesized, eg, 7-mers. Labeled probes are added to the immobilized probes of the corresponding unit array such that the addition of the ligase results in the covalent attachment of the probes and hybridization of the pair of immobilized and labeled probes adjacent to the target sequence. If the unit array contains two or more immobilized probes that are positive for a given nucleic acid species (as separate spots or within the same spot),
All corresponding labeled probes can be mixed and added to the same unit array. When testing mixtures of nucleic acid species, mixtures of labeled probes are even more important. One example of a complex mixture of nucleic acid species is mRN in one cell or tissue.
As.

According to one embodiment of the present invention, a unit array of immobilized probes allows the use of a solution of all possible immobilized probes and a relatively small number of labeled probes. When performing the multiplex labeling method, more complex solutions of labeled probes can be used. Preferred multiplexing methods can use different fluorescent dyes or molecular weight labels that can be separated by a mass spectrometer.

Alternatively, in accordance with a preferred embodiment of the present invention, relatively short immobilized probes can be selected that hybridize frequently to a large number of nucleic acid sequences. Such short probes are used in combination with a solution of labeled probes that can be made such that at least one labeled probe corresponds to each of the immobilized probes.
Preferred solutions are those in which none of the labeled probes corresponds to more than one immobilized probe.

Example 27 Interrogation of HIV Virus Segments Using All Possible 10-mers Format III In this example of SBH, all possible linked 5-mers (1
An array of 024 possible pentamers) was made on a nylon membrane (eg, GeneScreen). 5′-TTTTTT-NNN-3 ′ (N = all four bases A, C, G, T, at this stage of the reaction, adding an equal molar amount of all four bases) A conjugated 5-mer oligonucleotide is synthesized. Spot these oligonucleotides accurately on a nylon membrane,
The oligonucleotide was immobilized by drying the spot and treating the spot dried with UV light.

The subarrays in the array are separated from each other by a physical barrier, such as a hydrophobic strip, that allows each subarray to hybridize without being cross-contaminated by adjacent subarrays. In a preferred embodiment,
The hydrophobic strip is made from a silicone solution (eg, household silicone paste and seal paste) in a suitable solvent (such solvents are well known in the art). This solution of silicone grease is applied between the subarrays to form a line that acts as a hydrophobic strip separating cells after evaporation.

In this format III example, the free or solution (unbound) 5-crs is replaced by the 5′-NN-3 ′ (N = the 3 ′ end of all four bases A, C, G, T) Was synthesized using In this embodiment, the free 5-mer and the bound 5-mer are combined to allow all possible 10-mer sequences of less than 20 kb to be sequenced.
Make -mer. The 20 kb double strand is denatured to 40 kb single strand DNA. This 40 kb ss DNA hybridizes to about 4% of all possible 10-mers. Free or solution (unbound) for processing each subarray without loss of sequence information due to the low frequency of 10-mer binding sequences and known target sequences
It becomes possible to pool 5-mers. In a preferred embodiment, 16 probes are pooled for each array, and all possible 5-mers are free 5-mers for a total of 6
Expressed as pool 4. Thus, all possible 10-mers can be probed for DNA samples using 1024 subarrays (16 subarrays for each free 5-mer pool).

The target DNA in this embodiment represents two 600 bp segments of the HIV virus. These 600 bp segments have 60 overlapping 30-m
er (represented by a 30-mer every 20 nucleotides of each adjacent 30-m
er). The 30-mer pool is similar to target DNA that has been processed using techniques well known in the art to cut, digest, and / or random PCR the target DNA to create a random pool of very small fragments. ing.

The free 5-mer is labeled with a radioisotope, biotin, a fluorescent dye, etc., as described above in the Format III example above. The labeled free 5-mer is hybridized with the bound 5-mer to the target DNA and ligated. In a preferred embodiment,
Add 300-1000 units of ligase to the reaction. Hybridization conditions were performed according to the technique of the previous example. After ligation and removal of the target DNA and excess free probe, the array is assayed (using the techniques described in the Examples above) to determine the location of the labeled probe.

The known DNA sequence of the target in each subarray and the known free and bound 5-m
er predicts which binding 5-mer will link to the labeled free 5-mer in each subarray. The signal from 20 of these predicted dots is lost and the target D from the predicted sequence is lost.
Twenty new signals are obtained for each change in NA. The overlapping sequence of the binding 5-mer in these 10 new dots identifies which free labeled 5-mer binds in each new dot.

Using the method described above, probe the free labeled 5-mer, arrays of test HIV DNA sequences and pools with all possible 10-mers. This format II
Using Method I, we correctly identified the "wild-type" sequences of the tested segments, as well as some of the sequence "variants" introduced into these sequences.

[0344] In sequencing position embodiment of Example 28 repeated DNA sequences and sequenced using "spacer oligonucleotide" in format III method with improved repetitive DNA sequence in the target DNA. Spacer oligonucleotides of various lengths of the repetitive DNA sequence (the repetitive sequence is identified in a first round of SBH manipulation) are combined with a first known adjacent oligonucleotide and the other end of a second known or spacer. Hybridizes to the target DNA with a group of possible oligonucleotides flanked by (known well by the first round of SBH operation). When hybridizing to the target a spacer corresponding to the length of the repetitive DNA segment, two adjacent oligonucleotides can be ligated to the spacer. When a first known oligonucleotide is immobilized on a substrate and a second known or possible oligonucleotide is labeled, the appropriate length of the spacer will hybridize to the target DNA when the labeled second oligonucleotide is hybridized to the target DNA. A ligated product comprising two known or possible oligonucleotides is formed.

Example 29 Sequencing by Branching Point Using Format III SBH In one embodiment, branching points in target DNA are sequenced using a third set of oligonucleotides and a modified Format III method. First SBH
After the operation, when the array is compiled, some branch points can be identified. Hybridizing an oligonucleotide that partially overlaps one of the known sequences leading to the branch point, and then hybridizing to the target an additional oligonucleotide that is labeled and corresponds to one of the sequences that gives rise to the branch point. This will be solved. The appropriate oligonucleotide is the target DNA
When hybridized, the labeled oligonucleotide can be ligated to another. In a preferred embodiment, the first oligonucleotide that is offset from the branch point by one to several nucleotides is selected (as it is read into one of the branch sequences) and read from the first sequence into the sequence at the branch point. A second set of oligonucleotides (corresponding to the first oligonucleotide) corresponding to all possible branch sequences having an overlap of the sequence of the branch point every one to several nucleotides with the selection of the second oligonucleotide Select These oligonucleotides are hybridized to the target DNA, and only the third oligonucleotide having the appropriate branch sequence (which matches the branch sequence of the first oligonucleotide) is ligated to the first and second oligonucleotides. To produce

Example 30 Multiplexing Probes for Analyzing Target Nucleic Acid In this example, a set of probes is labeled with a different label such that each probe of the set is distinguished from the other probes in the set. Thus, the probe set can be contacted with the target nucleic acid in one hybridization reaction without losing any probe information. In a preferred embodiment, the different labels are different radioisotopes or different fluorescent labels or different EMLs
It is. These probe sets can be used in Format I, Format II or Format III SBH.

In the format ISBH, differently labeled probe sets are hybridized to target nucleic acids immobilized on a substrate under conditions that allow discrimination between perfect pairing and single base pair mismatches. Specific probes that bind to the target nucleic acid are identified by different labels, and perfect matching is determined, at least in part, by this binding information.

In format IISBH, target nucleic acids are labeled with different probes and hybridized to a probe array. Specific target nucleic acids that bind to the probe are identified by these different labels, and perfect matching is determined, at least in part, by this binding information.

In format III SBH, differently labeled probe sets and immobilized probes are hybridized to the target nucleic acid under conditions that allow discrimination between perfect pairing and single base pair mismatches. On the label, a labeled probe adjacent to the immobilized probe is bound to the immobilized probe, and these products are detected and distinguished by different labels.

In a preferred embodiment, the different labels are EMLs that can be detected on an electron capture mass spectrometer (EC-MS). EMLs are made from various backbone molecules where certain aromatic backbones are particularly preferred. See, for example, Xu et al. Chr
omatog. 764: 95-102 (1997). EML is bound in a reversible and stable manner, and when the probe is hybridized to the target nucleic acid,
EML is removed from the probe and identified by standard EC-MS (eg, EC-MS is performed by a gas chromatograph-mass spectrometer).

Example 31 Detection Format for Low Frequency Target Nucleic Acids III SBH has sufficient discriminating power to identify sequences differing by one nucleotide that are present in the sample at a ratio of 99: 1 of similar sequences.
Thus, Format III can be used to identify nucleic acids that are present at very low concentrations in nucleic acid samples, eg, samples from blood.

In one embodiment, the two sequences are cystic fibrosis sequences, which differ from each other by a deletion of three nucleotides. The two sequence probes were as follows: The probe that identifies the deletion from the wild type is fixed to the substrate,
The labeled proximity probe was common to both. Using these targets and probes, the deletion mutant, when present in 1 / 99th of the wild type, could be detected by format III SBH.

Example 32 Polaroid Instruments and Methods for Analyzing Target Nucleic Acids Any instrument for analyzing nucleic acids may be used to mix the two nucleic acid arrays and the nucleic acids of the two arrays until mixing is desired. It can be configured using a substance. The array of devices can be supported on a variety of substrates including, but not limited to, a nylon membrane, a nitrocellulose membrane, or other of the foregoing. In a preferred embodiment, one of the substrates is a membrane divided into multiple sectors by a hydrophobic strip, or a suitable support material with wells containing gels or sponges. In this embodiment, the probe is located on a sector of the membrane,
Alternatively, located in a well, gel, or sponge, a solution (with or without the target nucleic acid) is added to the membrane or well to solubilize the probe. Next, the solution containing the solubilized probe is brought into contact with the second nucleic acid array. The nucleic acid may be, but is not limited to, an oligonucleotide probe or a target nucleic acid, and the probe or the target nucleic acid may be labeled. Nucleic acids are radioisotopes,
It may be labeled with any label commonly used in the art, including, but not limited to, a fluorescent label or an electrophoretic label.

A substance that interferes with nucleic acid mixing can be placed between the two arrays in such a way that the nucleic acids of the two arrays mix when the substance is removed. This substance can be used in sheets, membranes,
Or it may be in the form of other barriers, and the material may be made of any material that prevents the mixing of diffusion.

This device can be used in Format I SBH as described below.
That is, a first array of instruments has target nucleic acids immobilized on a substrate, a second array of instruments has labeled nucleic acid probes, and upon removal, sends a responsive signal to the nucleic acids of the first array. Can be sent. The two arrays are optionally separated by a sheet of material that prevents the probes from contacting the target nucleic acid, and removal of the sheet allows the probes to send a responsive signal to the target. After appropriate incubation and (optional) washing steps, the target array can be "read" to see which probes have formed a perfect pair with the target. This reading can be automated or performed manually (eg, by eye reading using an autoradiogram). For Format II SBH, the following procedure will be similar to the previous one, except that the target will be labeled and the probe will be immobilized.

Alternatively, the device can be used in Format III SBH as described below. That is, two arrays of nucleic acid probes may be formed, one or both arrays of nucleic acid probes may be labeled, and one of the arrays may be immobilized on its substrate. The two arrays are separated by a sheet of material that keeps the probes from mixing. A Format II reaction is started by adding the target nucleic acid, removing the sheet and mixing the probes with each other and with the target. Probes that bind to adjacent sites on the target bind to each other (eg, by base-stacking interactions or covalent ligation of the backbone), and the results are read to determine which probes are binding to targets at adjacent sites. When a set of probes is immobilized on a substrate, the immobilized array can be read to determine which probes of the other array bind to the immobilized probes. As with the methods described above, this reading can be automated (eg, using an ELISA reader) or can be performed manually (eg, eye reading using an autoradiogram).

Example 33 Three-Dimensional Array of Probes In a preferred embodiment, oligonucleotide probes are immobilized on a three-dimensional array. A three-dimensional array includes multiple layers, such that each layer is analyzed separately from the other layers, at intervals, or all of the layers of the three-dimensional array are analyzed simultaneously. These three-dimensional arrays include, for example, arrays arranged on a substrate having multiple indentations in which the probes are located at different depths of the indentation (each level is made up of probes at the same indentation depth). Or an array arranged on a substrate with different depths of depressions where the probe is located at the bottom of the depression, the peak separating the depressions or the peak and depression indentations (each level being of a certain depth). Or made up of all probes); or an array disposed on a substrate that includes multiple sheets forming layers to form a three-dimensional array.

Methods for synthesizing these three-dimensional arrays are well known in the art and include the materials cited herein above as suitable as the support for the probe array. Also, other suitable materials, which can support oligonucleotide probes and are preferably flexible, can be used as substrates.

Example 34 Discriminating Sequences for Clustering cDNA Clones Multiple discrete nucleic acids from a cDNA library using standard pcr, SGH sequence discriminating characterization and Sanger sequencing methods. The sequence was obtained. The library insert was amplified using pcr using primers specific to the vector sequence adjacent to the insert. These samples were spotted on nylon membranes, probed with a suitable number of oligonucleotide probes, and the strength of the positive binding probe was determined by providing sequence signatures. Clones were clustered into similar or identical sequence signatures, and a signal representative of the clone was selected from each group for gel sequencing. The 5 'sequence of the amplified insert is deduced using reverse M13 sequencing primers in a typical Sanger sequencing protocol. The PCR product was purified and fluorescent dye terminator cycle sequencing was performed. 377 Applied Biosystems (Applied
Single-pass gel sequencing was performed using a Biosystems (ABI) sequencer. Most of the clones selected and sequenced by this method had sequences that differed from each other and only a few had the same sequence.

Example 35 High Throughput Production of Chips In a preferred embodiment, the apparatus for mass producing an array of probes is:
It includes a rotating drum or plate (eg, a microdrop injection head, etc.) coupled to an inkjet deposition apparatus, and a suitable robotic system (eg, an anorad gantry). A particularly preferred embodiment of this device is described with reference to FIGS.

The device comprises a cylinder (1) to which a suitable substrate has been fixed. The substrate may be made of any of the materials described above as suitable for the array of probes. In a preferred embodiment, the substrate is a flexible material and the array is made directly on the substrate. In another embodiment, a flexible substrate is secured to the cylinder and individual chips are secured on the substrate. The array is then made on each individual chip.

In a preferred embodiment, a physical barrier is provided on the substrate or chip to define an array of wells. The physical barrier may be provided on the substrate or chip by the device, or the physical barrier may be provided on the chip or substrate before being fixed to the cylinder (1). Next, one spot of the oligonucleotide probe is placed in each well, and the probes arranged in each well may have the same sequence, or the probe spotted in each well may have a different sequence. You may have. In a more preferred embodiment, the one or more probes spotted on each individual well in an array are different from the probes spotted on other wells of the array. Next, a sequencing chip containing a plurality of arrays can be assembled from these arrays.

After fixing the base or the base and the chip to the cylinder (1), the cylinder is rotated by a motor (not shown). The rotational speed of the cylinder is accurately measured by any method known in the art, such as using a stationary optical sensor and a light source that rotates with the cylinder. The dispensing device (3) moves along the arm (2) and applies probes and other reagents via a dispensing tip (8) using the exact rotational speed calculated as described above. It can be delivered to a precise location on a substrate or chip by methods known in the art. The distributor receives probes and reagents from a reservoir (6) via a supply line (7). The reservoir (6) contains all the probes and other reagents needed to make the array.

The dispensing device is shown in FIG. The dispensing device can have one or more dispensing tips (14 and 8). Each dispensing tip contains sample wells (13
) In the body (12), and the sample well (13) has a sample line (10).
Receive probes and other reagents through. The pressure line (11) is connected to the dispensing tip (14
And pressurize chamber (9) to psi sufficient to move probes and reagents through 8). Sample line (10), well (13) and distribution tip (14
And 8) must be washed between each change of probe or reagent between each change. An appropriate wash buffer is supplied to the sample well (13) via the sample line (10) or any dedicated wash line (not shown). Alternatively, optionally, part or all of the chamber (9) may be filled with a washing buffer. Then, if necessary, the wash buffer was added to the sample well (13) via the drain line (not shown) or via the sample line (10) and dispensing tips (14 and 8).
) And chamber (9).

Once the dispensing means has applied the probe to all appropriate locations in each array or chip, the substrate (with or without tip) is removed from the cylinder and a new substrate is fixed to the cylinder.

Example 36 Analysis of Target Nucleic Acid Using Probes Complexed with Discrete Particles In this embodiment, a target nucleic acid is used to form a complex (covalently or non-covalently) with a plurality of discrete particles. To find out. The discrete particles can be individually identified based on physical properties (or a combination of a plurality of physical properties), and the particles distinguished by the physical properties are complexed with different probes.
In a preferred embodiment, the probe is an oligonucleotide having a known sequence and length. Thus, the probe can be identified by the physical properties of the discrete particles. Probes suitable for this embodiment include all of the probes described in the previous section, including those with a shorter informative portion than the full length probe.

The physical properties of the discrete particles can be any properties known in the art that can divide the particles into several sets. For example, particles can be divided into several sets based on their size, fluorescence, absorbance, electromagnetic charge or weight. Alternatively, the particles can be labeled with a dye, radionuclide, or EML. Other suitable labels include ligands that can function as members that specifically bind to the labeled antibody, chemiluminescent substances, enzymes, and members that specifically bind to and bind to the labeled ligand. And the like.
A variety of easily used labels have been used in immunoassays. Still other labels include antigens, groups with specific reactivity, and components that are electrochemically detectable. Still other labels include any of the labels described in the previous section. These labels and properties can be measured quantitatively by methods known in the art, including, for example, those described in the previous section, and can distinguish particles based on signal intensity or signal type. (As one of the labels, for example, different dye strengths or different types of dyes may be applied to the particles). In a preferred embodiment, several physical properties are combined, and various combinations of these properties allow for the distinction of particles (e.g., combining 10 sizes and 10 colors into 100 particle groups, etc.). ).

The particle-probe allows the use of standard combinatorial methods, such as the ability to synthesize all possible 10-mers using about 2000 reaction vessels. 1024
An initial first set of reactions is performed to synthesize all possible 5-mers on 1024 differently labeled particles. The resulting probe-particles are mixed and divided into another set of 1024 reaction vessels. A second set of reactions is performed with these samples to synthesize all possible 5-mer extensions on probes in a pool of particles. This physical property identifies the first 5 nucleotides of each probe, and the reaction vessel identifies the second 5 nucleotides of all probes. Thus, all possible 10-mer probes are synthesized using 2048 reaction vessels. This method is simply modified to generate all possible n-mers for different probe lengths.

In a preferred embodiment, the particles are divided into several sets according to the fluorescence intensity of the particles. Since the particles in each set are prepared with varying intensities of the fluorescent labels, these particles have different fluorescent intensities. The fluorescence intensity of fluorescein is related to concentration over the range of 1: 300 to 1: 300,000 (Lockhart et al., 1986), with a linear relationship between 1: 3000 and 1: 300,000 (ie, fluorescein intensity is about 1 Linear over the range of ~ 300). In the linear range of detection, 256 sets of particles are labeled with fluorescein (eg, 3-259). With 256 sets of particles, all possible 4-mers can be linked to different sets of particles. By pooling the particles, all possible 5-mers have four pools of all possible 4-mers, and then the probes in each pool are only A, G, C or T It can be produced by stretching. Similarly, all possible 6-mers are all possible 4
-mer pools, in which each 4-mer pool is extended by one of the 16 possible 2-base permutations consisting of A, G, C and T. (For example, in the case of a 7-mer, 64 pools, 8 pools)
For example -256 pool for -mer).

The target nucleic acid is examined using a 5-mer probe (in 4 pools). The target nucleic acid is labeled with another fluorescent dye, or another different label (described above). The labeled target is mixed with the four pools, and the complementary probes in each pool will hybridize to the target nucleic acid. After detecting these hybridization complexes by methods known in the art, the hybridized positive probes are identified by detecting the fluorescence intensity of the particles. In many preferred embodiments, the mixture of probe-particles and target nucleic acid is supplied one particle at a time via a flow cytometer or other separation device, and the particle label and target are measured to determine Determine if the probe is complementary to the target nucleic acid.

In another embodiment, one set of free probes is labeled with another fluorescent dye or other label (as described above), and individual free probes are mixed with each pool of 5-mer probes (4 pools). This mixture is hybridized with the target nucleic acid. When the drug is added and the free probe is covalently linked to the 5-mer probe (see the description of the preferred drug above), the free probe is adjacent to the site on the target nucleic acid where the 5-mer probe binds. (A free probe site is a 5-m
er probe must be adjacent to the end). The particles are then assayed by methods known in the art to determine which particles have been covalently attached to the free probe (ie, the particles having a free probe label) and the 5-mer probe is identified by the fluorescence intensity of the particle. . In a most preferred embodiment, a mixture of probe-particles, free probe and target nucleic acid is provided one particle at a time via a flow cytometer and particle label and free probe label are measured to determine which probe is the target. Determine whether it is complementary to the nucleic acid.

In a preferred embodiment, a single device comprises all or most of the operations for analyzing a target nucleic acid using a probe-particle complex. The device has one or more reagent chambers in which the buffer and the labeled target nucleic acid are thoroughly mixed (target nucleic acid may be added manually or automatically). The mixture is aliquoted from the reagent chamber into a plurality of reaction chambers, each reaction chamber having one pool of probe-particle complexes. The probe-particle and the target nucleic acid react under conditions such that a probe complementary to the target nucleic acid is bound. Removing excess target nucleic acid (ie, unbound nucleic acid) from the reaction chamber (eg, by washing);
The particles bound to the target nucleic acid are identified by association of the target nucleic acid label with the particles, and the probe is identified by the physical properties of the particles. In a preferred embodiment, after removing excess target, the particles travel in a single row through the channel from the reaction chamber to the detector. As single particles pass through the detection devices, these devices measure the physical properties of the target label and the particles. In other preferred embodiments, before or after removing excess target, the particles can be sorted, for example, by size (eg, exclusion chromatography), by charge (eg, ion exchange chromatography), and / or
Or, by density-weight, divide into sets using one or a combination of these physical properties. These fractionated particles are then assayed by the detector.

As an alternative to this embodiment, the primary reagent chamber is supplied with buffer, target nucleic acid, pool of probe-particle complexes, and chemical or enzymatic ligation reagents. The components are thoroughly mixed and then aliquoted from the reagent chamber into multiple reaction chambers. Each reaction chamber has one labeled free probe. Alternatively, instead of adding the pool of probe particle complexes to the reagent chamber, the pool may be placed into the reaction chamber with the free probe. Further, a free probe can be added to the reagent chamber and a pool of probe particles can be added to the reaction chamber. The probe-particle, target nucleic acid, and free probe react under conditions such that the free probe and the particle-probe bind at adjacent sites on the target nucleic acid, thereby linking the free probe to the probe particle. You. Excess free probe (
That is, the unligated probe) and the target nucleic acid are removed from the reaction chamber (
Removed (by washing), the ligated probe is identified by the association of the free probe label with the particle, and the probe complexed with the particle is identified by the physical properties of the particle. In a preferred embodiment, after removing excess probe and target, the particles travel in a single row through the channel from the reaction chamber to the detector. When a single particle passes through the detection device, the device is free probe label covalently attached to the particle,
And the physical properties of the particles. In other preferred embodiments, before or after removing excess probe and target, the particles can be charged using their physical properties, eg, by size (eg, exclusion chromatography), by charge (eg, ion exchange chromatography). ) And / or density and weight. These fractionated particles are assayed by a detector.

In a preferred embodiment, there is a second set of reaction chambers in the device, where the pool of probe-particles is located in the second reaction chamber. The target and buffer are mixed in a reagent chamber and these are placed in a first reaction chamber containing a labeled free probe. The probe and target can be mixed and, optionally, the probe can be hybridized to the target. The labeled probe-target mixture is then sent to a second reaction chamber containing the probe-particle pool. The free probe and probe-particle hybridize to the target and the appropriate probe is ligated in the second reaction chamber. The linking agent may be added in the reagent chamber or in any of the reaction chambers, preferably, the linking agent is added in the second reaction chamber. The probe-particle hybridization products in the second reaction chamber are analyzed as described above.

In one embodiment, the target nucleic acid is not amplified (by PCR or by a vector such as the λ library) prior to analysis. Preferably, longer free and particle probes are used in this embodiment. This is because the sequence of the sample is more complex (ie to distinguish those that are positive against the background).

The probe-particle embodiments described in this example are useful for any of the foregoing applications, including but not limited to the diagnostic and sequencing applications described above.
In addition, embodiments of these probe particles can be modified by the above changes and modifications.

[0377] In the interaction this embodiment of the complementary polynucleotide in the presence of the agent that modifies the binding between Example 37 polynucleotide, the perfect match from mismatch in binding complementary polynucleotides Discrimination is adjusted by the addition of one or more drugs. In a preferred embodiment, the complementary polynucleotides are a target polynucleotide and a polynucleotide probe. The discrimination of a perfect match from a mismatch can be adjusted by the addition of the drug. The drug is a trialkylammonium salt (eg, TMAC
Ricelli et al., Nucl. Acids Res. 21: 3785-3788 (1993)), salts such as sodium chloride, phosphate and borate, and organic solvents such as formamide, glycol, dimethylsulfoxide, and dimethylformamide. , Urine, guanidinium, betaine and other amino acid analogs (Henke et al., Nucl. Acids Res. 19: 3957-3958 (1997
Rees et al., Biochemistry 32: 137-144 (1993)), polyamines such as spermidine and spermine (Thomas et al., Nucl. Acids Res. 25: 2396-2402 (1997)), or the negative charge of the phosphate backbone. Other positively charged molecules to be incorporated, surfactants such as sodium dodecyl sulfate, and sodium lauryl sarcosinate;
Minor groove binders, positively charged polypeptides, and intercalating agents such as acridine, ethidium bromide, and anthracene. In a preferred embodiment, a mixture of agents is added to the hybridization reaction to regulate the discrimination of a perfect match from a mismatch. Some of these agents distinguish by reducing the entropy of lysis between the two complementary strands.

In a preferred embodiment, the discrimination of a perfect match from a mismatch is improved by one or more agents. For example, formamide (a commonly used denaturant) was found to selectively destabilize (mispairing): (perfect pairing) in Format III reactions. As described above, the format III reaction was started, then varying amounts of formamide were added (0%, 10%, 20%, 30%, 40% and 50%). 0
In%, a perfect match signal was detected and the background (mismatch) was high. At 10% formamide, there was a good perfect match signal and the background / mismatch signal was reduced. At 20% formamide, the perfect match signal was reduced (albeit detectable) and the background / mismatch signal was abolished.
Between 30% and 50% formamide, there was no perfect match signal and no background / mismatch signal.

[0379] In another embodiment, using a drug to lower the T _m of a complementary polynucleotide of the pair. In a more preferred embodiment, to lower the T _m of a complementary polynucleotide of the pair using a mixture of agents. The drug can alter the Tm in various ways. Without limiting the invention, two examples include (1) agents that break hydrogen bonds between bases of two complementary polynucleotides (Goodman, Proc. N
at'l Acad. Sci. 94: 10493-10495 (1997); Moran et al., Proc. Nat'l Acad. Sci.
94: 10506-10511 (1997); Nguyen et al., Nucl. Acids Res. 25: 3059-3065 (1997)) and (2) neutralizing or neutralizing the negative charge of the phosphate in the sugar phosphate backbone of the polynucleotide. Drugs that block (Thomas et al., Nucl. Acids Res. 25: 2396-2402 (1997)). By enhancing or weakening (1) and / or (2), it is possible to regulate the _{Tm of} a pair of complementary polynucleotides.

In the most preferred embodiments, one or more agents are added to reduce the binding energy of GC base pairs, enhance the binding energy of AT base pairs, or both. In a preferred embodiment, one or more agents are added to make the binding energy from the AT base pair approximately equal to the binding energy of the GC base pair. Thus, the binding energy between these complementary polynucleotides depends entirely on the length. The binding energy of these complementary polynucleotides can be increased by adding agents that neutralize or mask the negative charge of the phosphate groups in the polynucleotide backbone.

The present invention is not intended to limit its scope by the illustrated embodiments, which are intended to illustrate aspects of the present invention and provide functionally equivalent compositions. And methods are within the scope of the present invention. Indeed, various modifications and changes will occur to those skilled in the art in practicing the present invention in light of the preferred embodiments herein. Accordingly, the scope of the invention is limited only by the claims. All references cited in the body of the specification are hereby incorporated by reference in their entirety.

[Brief description of the drawings]

FIG. 1 is a plan view of an apparatus for a mass-produced probe array.

FIG. 2 is a side view of an apparatus for a mass-produced probe array.

FIG. 3 is a cross-sectional side view of a dispensing unit of a device for mass-produced probe arrays.

──────────────────────────────────────────────────続 き Continued on the front page (31) Priority claim number 08/959, 365 (32) Priority date October 28, 1997 (Oct. 28, 1997) (33) Priority claim country United States (US) ( 81) Designated countries EP (AT, BE, CH, CY, DE, DK, ES, FI, FR, GB, GR, IE, IT, LU, MC, NL, PT, SE), OA (BF, BJ, CF, CG, CI, CM, GA, GN, GW, ML, MR, NE, SN, TD, TG), AP (GH, GM, KE, LS, MW, SD, SZ, UG, ZW), EA (AM, AZ, BY, KG, KZ, MD, RU, TJ, TM), AL, AM, AU, AZ, BA, BB, BG, BR, BY, CA, CN, CU, CZ, EE, GE, HR, H U, ID, IL, IS, JP, KG, KP, KR, KZ, LC, LK, LR, LT, LV, MD, MG, MK, MN, MX, NO, NZ, PL, RO, RU, SG , SI, SK, SL, TJ, TM, TR, TT, UA, US, UZ, VN, YU 4B024 AA01 AA11 AA20 CA01 CA09 CA11 HA11 HA14 HA19 4B063 QA01 QA13 QA17 QA18 QA19 QQ06 QQ07 QQ08 QQ10 QQ42 QQ52 QR31 QR41 QR48 QR49 QR50 QR51 QR56 QR62 QR66 QR84 QS02 QS25 QS34 Q02X13 AF01

Claims (64)

[Claims] Claims: 1. An array of multiple oligonucleotide probes, comprising a substrate and a material forming a physical barrier, forming a grid having a number of wells disposed on the substrate. Wherein the plurality of oligonucleotide probes are arranged to form an array in a number of wells, each well comprising one spot of the probe immobilized on the substrate. 2. The array of claim 1, wherein each individual spot has a probe with a different sequence than other probes in other spots of the array. 3. The array of claim 2, wherein the probes in each individual spot have the same sequence. 4. The array of claim 1, wherein the distance between the center of one spot and the center of an adjacent spot is at least 325 μm. 5. A sequencing chip comprising a plurality of arrays according to claim 1. 6. A sequencing chip comprising a plurality of arrays according to claim 2. 7. A sequencing chip comprising a plurality of arrays according to claim 3. 8. The array according to claim 1, wherein the oligonucleotide probe comprises an information providing moiety and a reactive group, and the probe is bound to the substrate by the reactive group. 9. The array of claim 8, wherein the oligonucleotide probe further comprises at least one randomized position. 10. The oligonucleotide probe further comprises a spacer.
An array according to claim 9. 11. An array of multiple oligonucleotide probes, comprising a substrate having multiple planes, wherein the multiple probes are immobilized on multiple planes of the substrate. 12. The array according to claim 11, wherein each plane can be analyzed individually. 13. The array according to claim 11, wherein multiple planes can be analyzed simultaneously. 14. The array of claim 11, wherein the oligonucleotide probe comprises an information providing moiety and a reactive group, wherein the probe is bound to the substrate by the reactive group. 15. The array of claim 14, wherein the oligonucleotide probe further comprises at least one randomized location. 16. The oligonucleotide probe further comprises a spacer.
An array according to claim 15. 17. A method for analyzing a target nucleic acid, comprising contacting the target nucleic acid with a number of oligonucleotide probes, provided that the probes are complexed with a number of different discrete particles that can be distinguished from each other based on physical properties. Different probes are attached to each type of discrete particle), detecting the probes that are complementary to the target nucleic acid, and analyzing the target nucleic acid from the set of complementary probes. Including the above method. 18. The method of claim 17, further comprising the step of separating the discrete particles to obtain a fraction, wherein the discrete particles are separated based on physical properties. 19. The method of claim 18, wherein the set of complementary probes has at least two overlapping probes. 20. The method according to claim 18, wherein the sequences of the target nucleic acids are combined in an analysis step. 21. The method of claim 18, wherein the complementary probes are identified by physical properties of the discrete particles. 22. The method of claim 18, wherein the physical property is associated with a molecule selected from the group consisting of a dye, a radionucleotide, an EML, and a fluorescent molecule. 23. The method of claim 18, wherein the physical property is selected from the group consisting of size, charge, absorbance, and weight. 24. The physical property is associated with the strength of the physical property.
3. The method according to 3. 25. The method of claim 2, wherein the physical properties are associated with a number of different molecules.
3. The method according to 3. 26. The method of claim 18, wherein the information providing portion of the probe is shorter than the full length probe. 27. The method of claim 18, wherein the target nucleic acid comprises a label and the complementary probe is detected by a label on the target nucleic acid. 28. The detecting step, wherein the individual particles are passed through a detector.
18. The method of claim 17, performed on individual discrete particles. 29. The method of claim 28, wherein the set of complementary probes has at least two overlapping probes. 30. The method of claim 28, wherein the sequences of the target nucleic acids are combined in an analysis step. 31. The method of claim 28, wherein the complementary probes are identified by physical properties of the discrete particles. 32. The method of claim 28, wherein the physical property is associated with a molecule selected from the group consisting of a dye, a radionucleotide, an EML, and a fluorescent molecule. 33. The method of claim 28, wherein the physical property is selected from the group consisting of size, charge, absorbance, and weight. 34. The physical property is associated with the strength of the physical property.
3. The method according to 3. 35. The physical property associated with a number of different molecules.
3. The method according to 3. 36. The method of claim 28, wherein the information providing portion of the probe is shorter than the full length probe. 37. The method of claim 28, wherein the target nucleic acid comprises a label and the complementary probe is detected by a label on the target nucleic acid. 38. Contacting a target nucleic acid with a number of free oligonucleotides and binding a complementary free probe bound to a site on the target nucleic acid to a site on the target nucleic acid adjacent to the site to which the free probe was bound 29. The method of claim 28, further comprising the step of covalently binding to a complementary probe of the discrete particle, wherein the detecting identifies a free probe covalently bound to the probe of the discrete particle. 39. The method of claim 38, wherein the set of complementary covalent probes has at least two overlapping covalent probes. 40. The method of claim 38, wherein the sequences of the target nucleic acids are combined in an analysis step. 41. The method of claim 38, wherein the probe complexed with the discrete particles is identified by the physical properties of the discrete particles. 42. The method of claim 38, further comprising separating the discrete particles to obtain a fraction, wherein the discrete particles are separated based on physical properties. 43. Separating the discrete particles with a flow cytometer to obtain a fraction,
43. The method of claim 42. 44. The method of claim 38, wherein the physical property is associated with a molecule selected from the group consisting of a dye, a radionucleotide, an EML, and a fluorescent molecule. 45. The method of claim 38, wherein the physical property is selected from the group consisting of size, charge, absorbance, and weight. 46. The physical property is related to the strength of the physical property.
5. The method according to 5. 47. The method of claim 4, wherein the physical properties are associated with a number of different molecules.
5. The method according to 5. 48. The method of claim 38, wherein the informative portion of the free probe is shorter than the full length probe. 49. The method of claim 38, wherein the information providing portion of the probe complexed with the discrete particles is shorter than the full length probe. 50. The method of claim 38, wherein the informative portion of the free probe and the informative portion of the probe complexed with the discrete particles are shorter than the full length probe. 51. A method for analyzing a target nucleic acid, comprising the steps of: combining a target nucleic acid with a probe under a condition enabling complete discrimination between a perfect match and a mismatch;
Wherein an agent is added to improve the discrimination between perfect and mispaired) and detecting whether the probe is complementary to the target nucleic acid. Method. 52. The drug may be a salt, organic solvent, urea, guanidinium, amino acid analog, polyamine, other positively charged molecule that neutralizes the negative charge of the phosphate backbone, surfactant, minor groove / major groove binder. 52. The method of claim 51, wherein the method is selected from the group consisting of: a positively charged polypeptide and an intercalating agent. 53. The salt according to claim 53, wherein the salt is a trialkylammonium salt, sodium chloride,
53. The method of claim 52, wherein the method is selected from the group consisting of phosphate and borate. 54. The method of claim 52, wherein said organic solvent is selected from the group consisting of formamide, glycol, dimethylsulfoxide and dimethylformamide. 55. The method of claim 52, wherein said amino acid analog is betaine. 56. The method of claim 52, wherein said polyamine is selected from the group consisting of spermidine and spermine. 57. The method of claim 52, wherein said surfactant is selected from the group consisting of sodium dodecyl sulfate and sodium lauryl sarcosinate. 58. The method of claim 52, wherein said intercalating agent is selected from the group consisting of acridine, ethidium bromide and anthracin. 59. The method of claim 51, wherein a plurality of drugs are added. 60. The drug may be a salt, organic solvent, urea, guanidinium, amino acid analog, polyamine, other positively charged molecule that neutralizes the negative charge of the phosphate backbone, surfactant, minor groove / major groove binder. 60. The method of claim 59, wherein the method is selected from the group consisting of: a positively charged polypeptide and an intercalating agent. 61. A method for analyzing a target nucleic acid, comprising: providing an array of a number of immobilized oligonucleotide probes; providing an array of a number of labeled oligonucleotide probes; The probe is used under conditions that enable the discrimination between a probe that forms a perfect pair with the target nucleic acid and a probe that binds to the target nucleic acid by a single base mismatch (here, a perfect match and a single base mismatch). A label that hybridizes the immobilized probe bound to a site on the target nucleic acid to a site on the target nucleic acid adjacent to the site to which the immobilized probe is bound. The above method comprising the steps of: covalently binding to a probe; and identifying immobilized and labeled probes covalently bound. 62. The drug may be a salt, organic solvent, urea, guanidinium, amino acid analog, polyamine, other positively charged molecule that neutralizes the negative charge of the phosphate backbone, surfactant, minor groove / major groove binder. 62. The method of claim 61, wherein the method is selected from the group consisting of: a positively charged polypeptide and an intercalating agent. 63. The method of claim 61, wherein said agent is formamide. 64. The method of claim 61, wherein a plurality of drugs are added.