Large-scale biomolecular analysis using sequence tags

Abstract

The present invention relates to performing sequence-based analysis of nucleic acid populations by multiplex amplification of one or more sequence tags, attaching the sequence tags to target nucleic acids and / or copies thereof, and then performing high-throughput sequencing of the amplified products. In some embodiments, the present invention includes the following consecutive steps: extending primers, removing unextended primers, and adding new primers for amplification (e.g., by PCR) or for additional primer extension. Some embodiments of the present invention relate to performing minimal residual disease (MRD) analysis on patients receiving cancer treatment. Introducing sequence tags into sequence reads provides an effective method for determining clonal types, while providing a convenient method for detecting carryover contamination from other samples from the same patient or carryover contamination from samples from other patients who have been tested in the same laboratory.

Description

[0001] This application is a divisional application of the invention patent application with application date of June 30, 2014, application number 2014800481194, and invention name “Large-Scale Biomolecule Analysis Using Sequence Tags”.

[0002] Cross-references

[0003] This patent application claims priority to U.S. Provisional Patent Application No. 61 / 841,878, filed on July 1, 2013, and U.S. Provisional Patent Application No. 62 / 001,580, filed on May 21, 2014, both of which are incorporated herein by reference in their entirety.

[0004] Sequence Listing

[0005] This application contains a sequence listing, which is submitted in ASCII format via EFS-Web and is hereby incorporated by reference in its entirety. The ASCII copy was created on June 27, 2014, is named 848US00-SL-ST25.txt, and has a file size of 2 kilobytes. No new objects have been added. Background Art

[0006] As the speed and convenience of large-scale DNA sequencing increase and its cost per base decreases, the application of large-scale DNA sequencing in diagnosis and prognosis has expanded rapidly, for example, Ding et al, Nature, 481(7382):506-510 (2012); Chiu et al, Brit. Med. J., 342:c7401 (2011); Ku et al, Annals of Neurology, 71(1):5-14 (2012), etc. In particular, nucleic acid profiles encoding immune molecules (such as T cell or B cell receptors, or components thereof) contain a wealth of information about the health status or disease of an organism, and thus such profiles have been proposed as diagnostic or prognostic indicators for a variety of conditions, e.g., Faham and Willis, US patents 8,236,503 and 8,628927; Freeman et al, Genome Research, 19: 1817-1824 (2009); Han et al, J. Immunol., 182(1001): 42.6 (2009); Boyd et al, J. Immunol., 182(1001): 42.6 (2009); al, Sci. Transl Med., 1(12): 12ra23 (2009) (Boyd et al., Science Translational Medicine, 2009, Vol. 1, No. 12, pp. 12-23); He et al, Oncotarget (March 8, 2011) (He et al., Oncotarget, March 8, 2011).

[0007] For example, many treated cancer patients often have minimal residual disease (MRD) associated with cancer. That is, even if the patient's disease can be completely relieved by clinical measures in response to treatment, a small part of the cancer cells may not be eliminated and remain for one reason or another. For the continued treatment of patients, the type and size of this residual population are important prognostic factors, such as Campana, Hematol. Oncol. Clin. North Am., 23 (5): 1083-1098 (2009) (Campana, "North American Hematology and Oncology Clinical", 2009, Vol. 23, No. 5, pp. 1083-1098); Buccisano et al, Blood, 119 (2): 332-341 (2012) (Buccisano et al., "Blood", 2012, Vol. 119, No. 2, pp. 332-341). Therefore, several techniques have been developed for assessing such populations, including those based on flow cytometry, in situ hybridization, cytogenetics, nucleic acid marker amplification, etc., such as Buccisano et al, Current Opinion in Oncology, 21: 582-588 (2009); van Dongen et al, Leukemia, 17(12): 2257-2317 (2003), etc. Amplification of recombinant nucleic acids encoding immune receptor segments (i.e., clonotypes) from T cells and / or B cells is particularly useful for assessing leukemia and lymphoma MRD, because such clonotypes often have unique sequences that can be used as molecular signatures for the associated cancer cells of these clonotypes. Because this type of amplification is highly multiplexed and difficult to establish, this type of determination is usually carried out partially by amplifying and sequencing the nucleic acid encoding a single receptor chain. Along with the increase in the scale of multiple amplification, several problems have been encountered, including: the probability of false amplification caused by false hybridization increases, the formation of primer dimers, the amplification rate of change causes sequence characterization bias, etc., such as Elnifro et al, Clinical Microbiology Reviews, 13 (4): 559-570 (2000) (Elnifro et al., "Clinical Microbiology Reviews", 2000, Vol. 13, No. 4, pp. 559-570). In addition, no matter for sequence analysis, sample tracking, contamination detection, or such work, the similarity of the target sequence and the incorporation of the sequence tag into the amplified sequence will aggravate the above-mentioned difficulties related to large-scale amplification.These difficulties have hampered the development of large-scale single-reaction amplification of multiple immune receptor chains, which would be highly advantageous in reducing the number of separate assays required to sequence nucleic acid sequences associated with microlesions.

[0008] In view of the above, it would be highly advantageous if more efficient methods existed for evaluating selected nucleic acids, such as exons of oncogenes or clonotypes encoding immune receptor chain sets, in a single reaction. Summary of the invention

[0009] The present invention relates to methods for large-scale amplification of a population of target polynucleotides (e.g., recombinant nucleic acids encoding immune receptor chains) in a single reaction, particularly by polymerase chain reaction (PCR), followed by large-scale DNA sequencing for identification. The present invention includes the application of the above methods to monitoring minimal residual lesions of cancer. The present invention is exemplified in a number of specific implementations and applications, some of which are summarized below and are presented throughout the specification.

[0010] In some embodiments, the present invention relates to a method for generating a nucleic acid spectrum, wherein the nucleic acid encodes a population of biological molecules (such as immune receptor molecules) of interest. In one aspect, the method of the present invention comprises: attaching a sequence tag to a selected nucleic acid population in a sample to form a tag-nucleic acid conjugate; amplifying the tag-nucleic acid conjugate; and sequencing the amplified tag-nucleic acid conjugate to provide sequence reads, each sequence read comprising a tag sequence and a nucleic acid sequence, and generating a nucleic acid spectrum for these sequences. In some embodiments, the attachment of the sequence tag is achieved by one or more consecutive primer extension and primer removal steps, and the resulting product can then be further amplified without bias by universal forward primers and reverse primers. In some embodiments, the present invention relates to methods for detecting and determining contamination (such as carryover contamination) in a sample caused by substances originating from different samples.

[0011] In one embodiment, such a method for detecting contamination in an individual undergoing minimal residual disease monitoring may include the following steps: (a) obtaining a tissue sample from the individual; (b) attaching a sequence tag to an oncogene molecule or a recombinant nucleic acid to form a tag-nucleic acid conjugate, wherein the sequence tag to which at least one nucleic acid or a copy thereof is attached is different, and wherein the oncogene molecule is unique to the individual's cancer; (c) amplifying the tag-nucleic acid conjugate; (d) sequencing a sample of the tag-nucleic acid conjugate to provide sequence reads having an error rate and comprising a tag sequence and an oncogene sequence or a recombinant nucleic acid sequence; (e) comparing the tag sequence with tag sequences determined separately from other tissue samples; and (f) determining the presence or absence of contamination and / or the extent of contamination by the identity of one or more tag sequences with any tag sequences determined separately from other tissue samples.

[0012] In another aspect, the invention relates to a method as described above for generating a clonotype profile based on at least two chains of a B-cell receptor, the method comprising amplifying a target nucleic acid encoding two or more chains of a B-cell receptor in a single reaction. In another aspect, such a method is used for monitoring minimal residual disease in B-cell cancer.

[0013] In another aspect, the present invention relates to a method for generating a clonotype profile based on at least two chains of a T cell receptor as described above, the method comprising amplifying a target nucleic acid encoding two or more chains of a T cell receptor in a single reaction. In another aspect, such a method is used to monitor minimal residual disease in T cell cancer.

[0014] These and other aspects of the present invention are exemplified in a number of illustrated implementations and applications, some of which are illustrated and characterized in the claims section below. However, the above summary is not intended to describe each illustrated embodiment or every implementation of the present invention.

[0015] The present invention also relates to the following items:

[0016] Item 1. A method for generating a clonal profile from a plurality of immune cell receptor chains, the method comprising the following steps:

[0017] (a) mixing a first set of primers with a sample of recombinant nucleic acids from B cells and / or T cells and / or cell-free DNA under primer extension conditions in a reaction mixture, wherein each primer in the first set has a receptor-specific portion, such that the receptor-specific portion anneals with a different recombinant nucleic acid at a predetermined position and then extends to form a first extension product, and wherein each primer in the first set has a 5' non-complementary end comprising a first primer binding site;

[0018] (b) removing the unextended primers in the first set from the reaction mixture;

[0019] (c) adding a second set of primers to the reaction mixture under primer extension conditions, wherein each primer in the second set has a receptor-specific portion, such that the receptor-specific portion anneals with the first extension product at a predetermined position and has a 5' non-complementary end comprising a second primer binding site, the primers of the first set and / or the primers of the second set respectively comprise a sequence tag located between the receptor-specific portion and the first or second primer binding site, and wherein each primer in the second set is extended to form a second extension product, such that each second extension product comprises a recombinant nucleic acid encoding a portion of a T cell receptor chain or a B cell receptor chain;

[0020] (d) performing a polymerase chain reaction in the reaction mixture to form an amplicon, the polymerase chain reaction using a forward primer specific for the first primer binding site and a reverse primer specific for the second primer binding site; and

[0021] (e) sequencing the nucleic acid of the amplicon to form a clonotype profile of a plurality of T cell receptor chains and / or B cell receptor chains.

[0022] Item 2. A method according to Item 1, wherein the multiple T cell receptor chains and / or B cell receptor chains include TCRβ, TCRδ and TCRγ, and wherein the primers of the first group and the primers of the second group include primers located on the side of the recombinant nucleic acid region encoding the VDJ region of TCRβ and TCRδ and primers located on the side of the VJ region of TCRγ.

[0023] Item 3. The method according to Item 1, wherein the sequencing step comprises:

[0024] generating sequence reads having an error rate and each comprising a tag sequence and a recombined nucleic acid sequence;

[0025] aligning sequence reads having similar tag sequences to form a group of sequence reads having the same sequence tag;

[0026] Merging the groups of sequence reads to determine clonotypes, wherein whenever there is at least a 95% probability that the groups of sequence reads are different, the groups of sequence reads are merged into different recombined nucleic acid sequences; and

[0027] The clonotype profile is formed from the clonotypes.

[0028] Item 4. According to the method described in Item 1, the method also includes a second step of removing the second extension product after it is formed.

[0029] Item 5. The method according to Item 1, wherein after the first extension product is melted, the annealing and extension of the first group of primers are repeated.

[0030] Item 6. The method according to Item 1, wherein after the second extension product is melted, the annealing and extension of the second group of primers are repeated.

[0031] Item 7. A method according to Item 1, wherein the removal step comprises adding exonuclease I activity to the reaction mixture.

[0032] Item 8. A method for generating a clonal profile from a plurality of B cell receptor chains, the method comprising the following steps:

[0033] (a) mixing a first set of primers with a sample of recombinant nucleic acids from B cells and / or cell-free DNA under primer extension conditions in a reaction mixture, wherein each primer in the first set has a receptor-specific portion such that the receptor-specific portion anneals with a different recombinant nucleic acid at a predetermined position and then extends to form a first extension product, and wherein each primer in the first set has a 5' non-complementary end comprising a first primer binding site;

[0034] (b) removing the unextended primers in the first set from the reaction mixture;

[0035] (c) adding a second set of primers to the reaction mixture under primer extension conditions, wherein each primer in the second set has a receptor-specific portion, such that the receptor-specific portion anneals with the first extension product at a predetermined position and has a 5' non-complementary end comprising a second primer binding site, the primers of the first set and / or the primers of the second set respectively comprise a sequence tag located between the receptor-specific portion and the first or second primer binding site, and wherein each primer in the second set is extended to form a second extension product, such that each second extension product comprises a first primer binding site, a second primer binding site, at least one sequence tag, and a recombinant nucleic acid encoding a portion of a B cell receptor chain;

[0036] (d) performing a polymerase chain reaction in the reaction mixture to form an amplicon, the polymerase chain reaction using a forward primer specific for the first primer binding site and a reverse primer specific for the second primer binding site; and

[0037] (e) sequencing the nucleic acid of the amplicon to form a clonotype profile of a plurality of B cell receptor chains.

[0038] Item 9. A method according to Item 8, wherein the multiple B cell receptor chains include IgH and IgK, and wherein the primers of the first group and the primers of the second group include primers located on the sides of the recombinant nucleic acid region encoding the VDJ region of IgH, the DJ region of IgH and the VJ region of IgK.

[0039] Item 10. The method according to Item 9, wherein the primers of the first group or the second group include at least one set of nested primers, and the at least one set of nested primers is specific to multiple different primer binding sites in the V region of the IgH chain.

[0040] Item 11. The method according to Item 8, wherein the sequencing step comprises:

[0041] generating sequence reads having an error rate and each comprising a tag sequence and a recombined nucleic acid sequence;

[0042] aligning sequence reads having similar tag sequences to form a group of sequence reads having the same sequence tag;

[0043] Merging the groups of sequence reads to determine clonotypes, wherein whenever there is at least a 95% probability that the groups of sequence reads are different, the groups of sequence reads are merged into different recombined nucleic acid sequences; and

[0044] The clonotype profile is formed from the clonotypes.

[0045] Item 12. The method according to Item 8, further comprising a second step of removing the second extension product after it is formed.

[0046] Item 13. The method according to Item 8, wherein after the first extension product is melted, the annealing and extension of the first group of primers are repeated.

[0047] Item 14. The method according to Item 8, wherein after the second extension product is melted, the annealing and extension of the second group of primers are repeated.

[0048] Item 15. A method according to Item 8, wherein the removal step comprises adding exonuclease I activity to the reaction mixture.

[0049] Item 16. A method for determining a clonal profile of a recombinant nucleic acid encoding a plurality of immune receptor chains in a sample, the method comprising the following steps:

[0050] (a) obtaining a sample comprising T cells and / or B cells and / or cell-free DNA from an individual;

[0051] (b) attaching sequence tags to recombinant nucleic acid molecules of T cell receptor genes or immunoglobulin genes from the sample to form tag-nucleic acid conjugates, wherein at least one recombinant nucleic acid or a copy thereof has a different sequence tag attached;

[0052] (c) amplifying the tag-nucleic acid conjugate;

[0053] (d) sequencing a sample of the tag-nucleic acid conjugate to provide sequence reads, each of the sequence reads having an error rate and each comprising a tag sequence and a recombinant nucleic acid sequence;

[0054] (e) aligning sequence reads with similar tag sequences to form a group of sequence reads with the same sequence tag;

[0055] (f) merging the sequence reads of each group to determine the clonotype, wherein whenever there is at least a 95% probability that the sequence read groups are different, the sequence read groups are merged into different recombined nucleic acid sequences; and

[0056] (g) determining the clonotype profile of the sample by determining the levels of the clonotypes.

[0057] Item 17. The method according to Item 16, wherein the attaching and amplifying steps comprise:

[0058] In a reaction mixture, a first set of primers is mixed with the sample under primer extension conditions, wherein each primer in the first set has a receptor-specific portion, such that the receptor-specific portion anneals with a different recombinant nucleic acid at a predetermined position and then extends to form a first extension product, and wherein each primer in the first set has a 5' non-complementary end comprising a first primer binding site;

[0059] removing the unextended primers in the first set from the reaction mixture;

[0060] adding a second set of primers to the reaction mixture under primer extension conditions, wherein each primer in the second set has a receptor-specific portion such that the receptor-specific portion anneals with the first extension product at a predetermined position and has a 5' non-complementary end comprising a second primer binding site, the primers of the first set and / or the primers of the second set respectively comprise a sequence tag located between the receptor-specific portion and the first or second primer binding site, and wherein each primer in the second set is extended to form a second extension product such that each second extension product comprises a first primer binding site, a second primer binding site, at least one sequence tag, and a recombinant nucleic acid encoding a portion of a T cell receptor chain or a B cell receptor chain; and

[0061] A polymerase chain reaction is performed in the reaction mixture to form an amplicon, the polymerase chain reaction using a forward primer specific for the first primer binding site and a reverse primer specific for the second primer binding site.

[0062] Item 18. The method according to Item 17, further comprising a second step of removing the second extension product after it is formed.

[0063] Item 19. The method according to Item 17, wherein after the first extension product is melted, the annealing and extension of the first group of primers are repeated.

[0064] Item 20. The method according to Item 17, wherein after the second extension product is melted, the annealing and extension of the second set of primers are repeated.

[0065] Item 21. A method according to Item 17, wherein the removal step comprises adding exonuclease I activity to the reaction mixture.

[0066] Item 22. The method according to Item 17, wherein the attachment and amplification steps will generate a tag-nucleic acid conjugate, wherein at least one of the recombinant nucleic acids and its corresponding copies are attached to multiple different sequence tags.

[0067] Item 23. A method according to Item 22, wherein the recombinant nucleic acid includes recombinant nucleic acids encoding TCRβ, TCRδ and TCRγ chains, and wherein the primers of the first group and the primers of the second group include primers located on the side of the recombinant nucleic acid region encoding the VDJ region of TCRβ and TCRδ and primers located on the side of the VJ region of TCRγ.

[0068] Item 24. A method according to Item 22, wherein the recombinant nucleic acid includes recombinant nucleic acid encoding IgH and IgK, and wherein the primers of the first group and the primers of the second group include primers located on the sides of the recombinant nucleic acid region encoding the VDJ region of IgH, the DJ region of IgH and the VJ region of IgK.

[0069] Item 25. A method according to Item 16, wherein the merging step comprises merging the sequence read groups into different recombined nucleic acid sequences whenever there is at least a 99.9% probability that the sequence read groups are different.

[0070] Item 26. A method for monitoring minimal residual lesions of a lymphoproliferative disease, the method comprising the following steps:

[0071] (a) obtaining a sample containing T cells and / or B cells and / or cell-free DNA or RNA from an individual;

[0072] (b) attaching a sequence tag to each of a plurality of recombinant nucleic acids in the sample to form a tag-nucleic acid conjugate, wherein at least one recombinant nucleic acid or a copy thereof has a different sequence tag attached and is specific for the lymphoproliferative disease of the individual;

[0073] (c) amplifying the tag-nucleic acid conjugate;

[0074] (d) sequencing a sample of the tag-nucleic acid conjugate to provide sequence reads, each of the sequence reads comprising a tag sequence and a recombinant nucleic acid sequence;

[0075] (e) aligning sequence reads with similar tag sequences to form a group of sequence reads with the same sequence tag;

[0076] (f) merging the groups of recombined nucleic acid sequences to determine clonotypes, wherein whenever there is at least a 99.9% probability that the groups of recombined nucleic acid sequences are different, the groups of sequence reads are merged into different clonotypes; and

[0077] (g) detecting the presence, absence and / or presence level of a clonotype associated with the lymphoproliferative disease in the individual in the clonotype profile.

[0078] Item 27. The method according to Item 26, wherein the attaching and amplifying steps comprise:

[0079] In a reaction mixture, a first set of primers is mixed with a sample of recombinant nucleic acids from B cells and / or T cells and / or cell-free DNA under primer extension conditions, wherein each primer in the first set has a receptor-specific portion such that the receptor-specific portion anneals with a different recombinant nucleic acid at a predetermined position and then extends to form a first extension product, and wherein each primer in the first set has a 5' non-complementary end comprising a first primer binding site;

[0080] removing the unextended primers in the first set from the reaction mixture;

[0081] adding a second set of primers to the reaction mixture under primer extension conditions, wherein each primer in the second set has a receptor-specific portion such that the receptor-specific portion anneals with the first extension product at a predetermined position and has a 5' non-complementary end comprising a second primer binding site, the primers of the first set and / or the primers of the second set respectively comprise a sequence tag located between the receptor-specific portion and the first or second primer binding site, and wherein each primer in the second set is extended to form a second extension product such that each second extension product comprises a recombinant nucleic acid encoding a portion of a T cell receptor chain or a B cell receptor chain;

[0082] A polymerase chain reaction is performed in the reaction mixture to form an amplicon, the polymerase chain reaction using a forward primer specific for the first primer binding site and a reverse primer specific for the second primer binding site.

[0083] Item 28. A method according to Item 27, wherein the recombinant nucleic acid includes recombinant nucleic acids encoding TCRβ, TCRδ and TCRγ chains, and wherein the primers of the first group and the primers of the second group include primers located on the side of the recombinant nucleic acid region encoding the VDJ region of TCRβ and TCRδ and primers located on the side of the VJ region of TCRy.

[0084] Item 29. A method according to Item 27, wherein the recombinant nucleic acid includes recombinant nucleic acid encoding IgH and IgK, and wherein the primers of the first group and the primers of the second group include primers located on the sides of the recombinant nucleic acid region encoding the VDJ region of IgH, the DJ region of IgH and the VJ region of IgK.

[0085] Item 30. A method according to Item 29, wherein the primers of the first group or the second group include at least one set of nested primers, and the at least one set of nested primers is specific to multiple different primer binding sites in the V region of the IgH chain.

[0086] Item 31. The method according to Item 30, further comprising a second step of removing the second extension product after it is formed.

[0087] Item 32. The method according to Item 30, wherein after the first extension product is melted, the annealing and extension of the first group of primers are repeated.

[0088] Item 33. The method according to Item 3, wherein after the second extension product is melted, the annealing and extension of the second group of primers are repeated. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] The novel features of the present invention are particularly set forth in the appended claims. The features and advantages of the present invention may be better understood with reference to the following detailed description of exemplary embodiments and the accompanying drawings, in which the principles of the present invention are utilized, and which include:

[0090] Figures 1A to 1C Various embodiments of the invention are shown by way of illustration. Figure 1D Methods of generating templates for recombinant nucleic acids of predetermined length (with or without the use of sequence tags) are shown.

[0091] Figures 2A to 2G Various methods of attaching a unique sequence tag to substantially every target sequence in a sample are shown.

[0092] Figure 3A and Figure 3B One aspect of the invention is schematically shown for generating clonotype profiles from nucleic acid sequences encoding IgH chains.

[0093] Figure 4A The use of sequence tags in determining the clonotype sequence of a sequence read is shown. Figure 4B The use of sequence tags in embodiments where multiple different sequence tags are attached to the same target polynucleotide or copies thereof is shown.

[0094] Figure 5A The concept of distances between clonotypes and closely related clonotypes in sequence space is shown.

[0095] Figure 5B is a flow chart illustrating one embodiment of a method for distinguishing truly distinct clonotypes from clonotypes that differ only due to sequencing errors (which should be merged).

[0096] Figure 5C A form of a numerical function is shown that is used in one embodiment to determine whether to merge correlated clonotypes.

[0097] Figure 5D and Figure 5E The use of a sequence tree in a method for merging sequence reads is shown. DETAILED DESCRIPTION

[0098] Unless otherwise indicated, the implementation of the present invention may adopt conventional techniques and descriptions of molecular biology (including recombinant technology), bioinformatics, cell biology and biochemistry, which are within the technical scope of the art. Such conventional techniques include, but are not limited to, sampling and analysis of blood cells, nucleic acid sequencing and analysis, etc. Specific examples of suitable techniques can be obtained with reference to the examples below. However, other equivalent conventional methods can also be used, of course. Such conventional techniques and descriptions can be found in standard laboratory manuals, such as Genome Analysis: A Laboratory Manual Series (Vols. I-IV) ("Genome Analysis: Laboratory Manual Series" (I-IV Volumes)); PCR Primer: A Laboratory Manual ("PCR Primer: Laboratory Manual"); and Molecular Cloning: A Laboratory Manual ("Molecular Cloning: Laboratory Manual") (both published by Cold Spring Harbor Laboratory Press) and the like.

[0099] In one aspect, the present invention relates to a method for generating a clonal profile of multiple immune receptor chains by performing large-scale multiplex amplification of nucleic acids encoding such chains, followed by high-throughput sequencing of the amplified products (or amplicons). In some embodiments, the present invention overcomes the common defects of multiplex amplification by including some consecutive steps, including: extending primers, removing unextended or unincorporated primers, and adding new primers for amplification (e.g., by PCR) or for additional primer extension. These steps also enable the use of sequence tags, which otherwise would contribute to nonspecific amplification or false amplification. On the other hand, sequence tags are used in embodiments of clinical applications, particularly for minimal residual disease (MRD) analysis (e.g., for samples from patients undergoing cancer treatment). Sequence tags incorporated into sequence reads provide an effective means for determining clonal types, while also providing a convenient means for detecting carryover contamination by detecting the presence or absence of sequence tags by the aforementioned assay, and the sequence tags may be from samples of the same patient or from samples of different patients tested in the same laboratory. Of particular interest are methods that use a single amplification reaction to generate a sequence-based clonotype profile for a recombinant nucleic acid encoding multiple B cell receptor (BCR) chains, followed by a new generation of high-throughput sequencing. Also of particular interest are methods that use a single amplification reaction to generate a sequence-based clonotype profile for a recombinant nucleic acid encoding multiple T cell receptor (TCR) chains, followed by a new generation of high-throughput sequencing. The methods of the present invention may also be applied to other large-scale amplifications and sequencing of other nucleic acid groups of interest (including, for example, exomes of oncogenes). In these aspects, sequence tags are used to monitor carryover contamination and to more sensitively determine the nucleotide sequence of a target polynucleotide based on an error-prone sequencing method. In addition, in these aspects, a set of sequence tags (as discussed more fully below) is typically much larger than the number of target polynucleotides in a sample, and the sequence differences between the sequence tags attached to the target polynucleotides are large enough to effectively avoid the conversion of one tag sequence into another tag sequence due to sequencing errors.

[0100] An embodiment of the present invention is Figure 1AAs shown. In the reaction mixture, the primer 22 from the first group (each primer of the first group has a receptor-specific portion 16 and a 5' non-complementary portion 15 comprising a first primer binding site) is annealed and paired with one end of the target polynucleotide 10 (after the target polynucleotide 10 is melted), and the primer 24 from the second group (each primer of the second group has a receptor-specific portion 20 and a 5' non-complementary portion, and the 5' non-complementary portion comprises a sequence tag 14 and a second primer binding site 12) is annealed and paired with the other end of the target polynucleotide 10. In some embodiments, as shown below, the non-complementary portion 15 of the primer 22 may also include a sequence tag. In some cases, two shorter sequence tags may be more advantageous than a single longer sequence tag with equal diversity. Therefore, for example, two 8-poly random nucleotide sequence tags may be less likely to cause false starts, primer dimers, etc. than a single 16-poly random nucleotide sequence tag. The target polynucleotide 10 is typically a somatic recombinant nucleic acid from a T cell or a B cell that encodes a chain or a portion of a chain (e.g., a portion of an IgH chain or an IgK chain) of a T cell receptor (TCR) or a B cell receptor. Thus, in some embodiments, the receptor-specific portions of primers 22 and 24 may be specific to a V region sequence and a J region sequence, respectively, or, in other embodiments, vice versa.

[0101] In some embodiments, the target polynucleotide 10 may comprise a complex mixture of nucleic acids of a desired sequence profile, including but not limited to: recombinant nucleic acids encoding portions of immune receptor molecules; 16SrDNA of microbial communities; genes encoding proteins (such as enzymes) that are important industrially or medically, genes and / or metagenomic amplification of exons of humans or animals associated with specific diseases (such as cancer, infectious diseases, etc.). In embodiments involving recombinant nucleic acids encoding immune receptors, typically at least a portion of the V region, D region, or J region is present between the two binding positions of the first set of primers and the second set of primers. In some embodiments, there is at least a portion of the VDJ rearrangement of IgH, the DJ rearrangement of IgH, the VJ rearrangement of IgK, the VJ rearrangement of IgL, the VDJ rearrangement of TCRβ, the DJ rearrangement of TCRβ, the VJ rearrangement of TCRα, the VJ rearrangement of TCRγ, the VDJ rearrangement of TCRδ, or the VD rearrangement of TCRδ between the two binding positions of the first set of primers and the second set of primers. In some embodiments, there is at least a portion of VDJ rearrangement of IgH, DJ rearrangement of IgH, VJ rearrangement of IgK, or VJ rearrangement of IgL between the two binding positions of the first set of primers and the second set of primers. In some embodiments, there is at least a portion of VDJ rearrangement of TCRβ, DJ rearrangement of TCRβ, VJ rearrangement of TCRα, VJ rearrangement of TCRγ, VDJ rearrangement of TCRδ, or VD rearrangement of TCRδ between the two binding positions of the first set of primers and the second set of primers. In other embodiments, there is at least a portion of VDJ rearrangement of IgH, DJ rearrangement of IgH, and VJ rearrangement of IgK between the two binding positions of the first set of primers and the second set of primers. And, in other embodiments, there is at least a portion of VDJ rearrangement of TCRβ, VJ rearrangement of TCRγ, VDJ rearrangement of TCRδ, or VD rearrangement of TCRδ between the two binding positions of the first set of primers and the second set of primers. In some embodiments, at least a portion of the VDJ rearrangement comprises a complete D or NDN portion and a portion of the V segment and the J segment that satisfies the requirements for its identification. In some embodiments, at least a portion of the VDJ rearrangement comprises a segment of at least 50 nucleotides, the 50 nucleotide segment comprising a complete D or NDN portion and a portion of the V segment and the J segment. In some embodiments, at least a portion of the VDJ rearrangement comprises a segment of at least 70 nucleotides, the 70 nucleotide segment comprising a complete D or NDN portion and a portion of the V segment and the J segment.

[0102] In some embodiments, the first group includes one or more primers, each of which is specific to the J segment or the C segment. Primers from such a first group are annealed to their target sequences and then extended, after which the primers that are not extended in the first group are removed. Primers from the second group that are each specific to the V segment are annealed to their target sequences and then extended. In other embodiments, the first group includes primers that are each specific to the V segment, such primers of the first group are annealed to their target sequences and then extended, after which the primers that are not extended in the first group are removed; primers in the second group that are each specific to the J segment or the C segment are annealed to their target sequences and then extended. In alternative forms of these two embodiments, the first group and the second group may each include multiple primers, and each primer may be specific to different immune receptor segments.

[0103] Back to Figure 1AIn some embodiments, the first and second sets of primers 5 are extended by performing 1-10 or 2-10 or 3-10 or 4-10 or 5-10 cycles of melting, annealing and extension in alternative embodiments, followed by removal of the unextended primers from the reaction mixture using conventional techniques. In other embodiments, the first and second sets of primers 5 are extended by performing 2-5 or 3-5 or 4-5 cycles of melting, annealing and extension in alternative embodiments, followed by removal of the unextended primers from the reaction mixture using conventional techniques. In another embodiment, the first and second sets of primers are extended by performing two cycles of melting, annealing and extension. For example, unextended primers can be removed using exonuclease digestion, hybridization with complementary sequences on magnetic beads, size exclusion chromatography, commercially available spin columns (such as Qiagen QIAquick PCR Purification Kit), etc. In one embodiment, unextended or unincorporated primers are removed by, for example, digestion with exonuclease I. The double-stranded DNA 18 is an extended product 5, which has a universal first and second primer binding site at each end, to which forward and reverse primers having complementary sequences 6 and 11 can be added (in some embodiments) for later generation of clusters by bridge PCR. In some embodiments, the double-stranded DNA also has a sequence tag 19, and the forward or reverse primer can include a sample tag 2 for identifying or tracking the DNA 18, or associating it with a sample or patient. In some embodiments, the sequence tag 19 is essentially unique for each different recombinant nucleic acid in the sample. As will be described more fully below, the sequence tag 19 can be used to merge sequence reads into clonal types, and can also be used to detect and track sample contamination. The forward and reverse primers can also include primer binding sites 4 and 8 for performing (13) bridge PCR on (for example) a Genome Analyzer 17 (Illumina, San Diego) for certain sequencing protocols. In other embodiments where more than one extension is performed with primers comprising sequence tags, each different recombinant nucleic acid in the sample may have a copy with a different sequence tag attached; thus, for example, if Figure 1A The illustrated embodiment subjects the target polynucleotide to four separate cycles of melting, annealing, and extension, and if the sample contains recombinant nucleic acid S1, then when amplification (13) is performed using universal primers, the copies of S1 will have at most four different sequence tags. Thus, the sequence reads of S1 will have at most four different sequence tags. As will be described more fully below, in such embodiments, clonotypes can be determined by aligning sequence tags in combination with merging sequence reads within each subset defined by the universal sequence tags.

[0104] In another embodiment, at least two extension steps and two steps of removing unincorporated primers are performed before PCR with universal primers. Figure 1B As shown, primer 101 is annealed to one end of a target polynucleotide 100 (e.g., a recombinant nucleic acid encoding an immune receptor chain) and then extended, for example, using a DNA polymerase. Primers 101 may each include a receptor-specific portion 103 and a 5' non-complementary portion 105, which in turn includes a sequence tag 104 and a first primer binding site 102. After extension and removal of unincorporated primers 130 as described above, the following are added to the first extension product 109 in the reaction mixture: (a) primers 125, each of which includes a receptor-specific portion 106 and a 5' non-complementary portion 115 (which includes a primer binding site); and (b) primer 127, which includes a portion 108 specific to the first primer binding site 102 and the 5' non-complementary portion 117. After primers 125 and 127 are annealed to their primer binding sites, these primers are extended (107) to form a second extension product 118, and then the unextended primers are removed. A universal forward primer 112 and a reverse primer 110 are added to the second extension product 118 and PCR is performed (111), and then the resulting amplicon is sequenced (120). Figure 1A As described in the embodiment of , whenever more than one extension step is performed in the presence of a primer comprising a sequence tag (such as 101), copies of the same target polynucleotide 100 can be labeled with multiple different sequence tags.

[0105] Figure 1C Another embodiment is shown that explicitly shows the V region, D region, and J region. Primers 1212 (i.e., a first set of primers specific to V region 1226) are added to a recombinant nucleic acid 1200 encoding an immune receptor (e.g., a TCR) in a reaction mixture under primer annealing conditions. Each primer of the first set 1212 comprises a receptor-specific portion and a 5' non-complementary portion, which in turn optionally comprises a sequence tag and a first primer binding site (e.g., Figure 1B102, 103 and 104 shown). The first set of primers 1212 anneal to the V region 1226 of the recombinant nucleic acid 1200, and then the first set of primers 1212 are extended (1202) through the D region 1224 to at least the J region 1222, and optionally to the C region 1220, to form a first extension product 1216 comprising an optional sequence tag 1228 and a first primer binding site 1230. After removing the unextended primers of the first set 1212, a second set of primers 1240 are added to the reaction mixture under annealing conditions so that these primers anneal to their corresponding target J region 1222, and then these primers are extended (1204) to form second extension products 1232, each of which comprises a sequence tag 1236 (optional) and a second primer binding site 1234. The second extension product 1232 may include a single sequence tag, which is located, for example, near the V region 1226 as shown by sequence tag 1228, or near the J region 1222 as shown by sequence tag 1236; or the second extension product 1232 may include two sequence tags located at both positions. In one embodiment, the second extension product 1232 includes a single sequence tag 1228 adjacent to the V region 1226. In another embodiment, the second extension product 1232 includes a single sequence tag 1236 adjacent to the J region 1222. In some embodiments, sequence tags 1228 and / or 1236 are mosaic tags as described below. After removing the unextended primers of the second set 1240, a universal forward primer and a reverse primer specific to the first primer binding site 1230 and the second primer binding site 1234, respectively, are added, and then PCR is performed (1206). A sample of the resulting amplicons is sequenced (1208) to generate sequence reads for constructing clonotypes and clonotype profiles.

[0106] Figure 1D A method of generating a template of defined length and attaching one or two sequence tags thereto is shown. Figure 1DThe embodiment of the invention shows messenger RNA (mRNA) as a starting material, but the method can be applied to any of DNA or RNA samples. For mRNA 1300 containing VDJ region, one or more primers 1312 ("C primers") having specificity to C region 1308 are annealed and paired with mRNA 1300. Usually only a single C primer is used. Alternatively, one or more primers (having similar structures) having specificity to J region can be used. C primer 1312 includes a target-specific segment 1313, a sequence tag segment 1314, and a universal primer binding site 1315. A polymerase blocker 1310 is also annealed and paired with the target mRNA 1300, and the polymerase blocker can be an oligonucleotide having specificity to V region 1302. In some embodiments, as long as the polymerase used to extend primer 1312 has no chain displacement activity or 5'→3' exonuclease activity, and the oligonucleotide is not extendable (for example, it has a 3'-dideoxynucleotide), the blocker 1310 can be a natural oligonucleotide. Typically, blocker 1310 is an oligonucleotide analog with enhanced binding activity and nuclease resistance, such as an antisense compound. In some embodiments, the blocker 1310 may be a locked nucleic acid (LNA) or a peptide nucleic acid (PNA) or a bridge nucleic acid (BNA), which are disclosed in the following references: Wengel et al, US patents 6,794,499; 7,572,582 (Wengel et al, US Patents 6,794,499, 7,572,582); Vester et al, Biochemistry, 43 (42): 13233-13241 (2004) (Vester et al, Biochemistry, 2004, Vol. 43, No. 42, pp. 13233-13241), etc.; and Kazuyuki et al, Chem. Comm., 3765-3767 (2007) (Kazuyuki et al, Chemical Communications, 2007, pp. 3765-3767); Nielson et al, Chem. Comm., 3765-3767 (2007); al, Chem. Soc. Rev., 26: 73-78 (1997) (Nielson et al., "Chemical Society Reviews", 1997, Vol. 26, pp. 73-78), etc. The sequence of the blocker 1310 is selected so that the extension of the primer 1312 stops at a predetermined position on the V region 1302. In some embodiments, the blocker 1310 is designed so that only enough V region 1302 is copied during the extension step, so that the V region can be identified from the copied sequence. In some embodiments, since a common sequence that allows some mismatches to occur can be selected as long as the polymerase can be stopped, it is not necessary to obtain a blocker 1310 that is specific to each V region. The length of the blocker 1310 can vary greatly depending on the type of oligonucleotide or oligonucleotide analog used.In some embodiments, the length of blocker 1310 ranges from 10 to 25 monomers. In some embodiments, blocker 1310 can anneal to different positions on different V region sequences.

[0107] Back to Figure 1D , primer 1312 is extended to blocker 1310 to form a cDNA copy with a predetermined length, which is a portion of the VDJ region of target 1300. In some embodiments, the predetermined length (or in other words, the binding site of blocker 1310) is selected so that one or more sequence reads measured by the sequencing technology used in the method cover the desired portion of the VDJ region. After the extension is completed, the RNA template 1300 is digested (1325) using conventional techniques (for example, digested with ribonuclease (RNase) such as RNase H and / or RNase A) to obtain a single-stranded cDNA 1326. Terminal deoxynucleotidyl transferase (TdT) in conventional protocols is used to add a 3' single nucleotide tail, such as a poly-C tail, to the cDNA. For tailed cDNA 1331, adaptor 1336 has an overhang complementary to the single nucleotide tail of cDNA 1331, after which the adaptor is extended to produce double-stranded DNA 1340, which can be amplified by (for example) PCR (1337) and the resulting amplicon sequenced (1338).

[0108] A recombinant nucleic acid that has undergone a high-frequency mutation (such as a nucleic acid encoding IgH) can be amplified using a primer set that includes primers that bind to different primer binding sites on the same recombinant nucleic acid; that is, such a set may include primers that bind to one or more non-overlapping primer binding sites on the same recombinant nucleic acid encoding the receptor chain. Such a set may include one or both of a first set of primers and a second set of primers. In some embodiments, a recombinant nucleic acid that has undergone a high-frequency mutation is amplified using a first set of primers and a second set of primers, wherein at least one of the two groups includes primers that are specific to multiple non-overlapping primer binding sites, for example, one group may include multiple primers for each different V segment, each primer having specificity for binding to different non-overlapping primers on different V segments. An embodiment suitable for amplifying a recombinant nucleic acid that has undergone a high-frequency mutation is as follows: Figure 3A-3BAs shown, the nested set of primers is used to ensure that each recombinant nucleic acid in the sample is amplified under conditions such as somatic high frequency mutation, clonal evolution, etc. Under annealing conditions, the recombinant nucleic acid (such as a recombinant nucleic acid encoding an IgH molecule) is combined with the first nested set of primers 302 in the reaction mixture, in this example, the first nested set of primers includes primer groups 304, 306 and 308, and these primer groups are specific to different sites on the V region 316 of the recombinant nucleic acid 300. In this embodiment, the first nested set includes multiple primer groups, each primer group is specific to a different site or position in the V region, and different members in one group are specific to different variants of the site in the V region. In some embodiments, the multiple groups are in the range of 2 to 4; in other embodiments, the multiple is 2 or 3. In some embodiments, each primer of the first nested set 302 may have a unique sequence tag 314 and a first primer binding site 312 in the 5' non-complementary tail. The first nested set of primers 302 anneal to their target recombinant nucleic acid and extend through the D region 318 and at least a portion of the J region 320 to form a first amplicon 323, which includes three components 330, 332, and 334 corresponding to the three primer subsets 304, 306, and 308. Each member of the first amplicon 323 includes a sequence tag 324 and a primer binding site 326.

[0109] After removing (322) the unextended primers, a second nested set of primers 340 is added to the reaction mixture under annealing conditions. Figure 3AAs shown, the second nested set of primers 340 includes primer subsets 336 and 338, which anneal and pair at different non-overlapping positions on the J region 320 of the first amplicon 323 member. In some embodiments, the second nested set of primers may only include a single primer group. The second nested set of primers 340 is extended to form a second extension product 360, which includes subsets 350, 352 and 354, each of which includes two further subsets (sub-subsets) corresponding to primers 336 and 338. In some embodiments, the second nested set of primers 340 includes primers that are specific only to a single primer binding site, while the first nested set of primers 302 includes primers that are specific to at least two non-overlapping primer binding sites. After removing the unextended primer (342), a universal forward primer and a reverse primer can be added for PCR (356), and then a sample of the resulting amplicon can be sequenced (358). In various embodiments, the primers of the first nested set and the second nested set may both include sequence tags 339; the primers of the first nested set may include sequence tags while the primers of the second nested set may not include sequence tags; the primers of the second nested set may include sequence tags while the primers of the first nested set may not include sequence tags. In some embodiments, the primers of the first nested set are first extended, then the unextended primers are removed or destroyed, and then the primers of the second nested set are annealed and extended (e.g., Figure 3A-3B In other embodiments, the order of the annealing, extension, and removal steps is reversed; that is, the second nested set of primers is first extended, followed by removal or destruction of the unextended primers, and then the first nested set of primers is annealed and extended.

[0110] In some embodiments of the above method, more than one extension step (322 or 342) may be performed, for example, to attach sequence tags to a greater proportion of target polynucleotides in the sample. In such embodiments, more than one different sequence tag may be attached to the target polynucleotide and / or its copy. That is, a plurality of different sequence tags may be attached to the target polynucleotide and its progeny from an amplification reaction (such as PCR); therefore, copies of the original target polynucleotide may be labeled with more than one sequence tag. As will be more fully described below, such multiple sequence tags are still useful in tracking carryover contamination and allowing for more sensitive determination of the target polynucleotide sequence.

[0111] Some of the above embodiments can be performed by the following steps. For example, a method for generating a clonal profile from a plurality of or more T cell receptor chains may include the following steps: (a) in a reaction mixture, mixing a first set of primers with a recombinant nucleic acid sample from a T cell under primer extension conditions, wherein each primer of the first set has a receptor-specific portion, the receptor-specific portion having a certain length so that it anneals and pairs with a different recombinant nucleic acid at a predetermined position or site on the target recombinant nucleic acid, and then extends to form a first extension product, and wherein each primer of the first set has a 5' non-complementary end comprising a first primer binding site; (b) removing the unextended primers of the first set from the reaction mixture; (c) adding a second set of primers to the reaction mixture under primer extension conditions, wherein each primer of the second set has a receptor-specific portion so that the receptor-specific portion anneals and pairs with the first extension product at a predetermined position or site, and has a 5' non-complementary end comprising a second primer binding site, and the first set of primers The primers and / or the second set of primers comprise a sequence tag located between the receptor-specific portion and the first or second primer binding site, respectively, and wherein each primer of the second set is extended to form a second extension product, whereby each second extension product comprises the first primer binding site, the second primer binding site, at least one sequence tag, and any of: (i) a portion of a Vβ segment and a portion of a Jβ segment of a T cell receptor chain, (ii) a portion of a Vγ segment and a portion of a Jδ segment of a T cell receptor chain, or (iii) a portion of a Vγ segment and a portion of a Jγ segment of a T cell receptor chain; (d) performing a polymerase chain reaction in a reaction mixture to form an amplicon, the polymerase chain reaction using a forward primer specific for the first primer binding site and a reverse primer specific for the second primer binding site; and (e) sequencing the nucleic acid of the amplicon to form a clonal profile of a plurality of T cell receptor chains. As used herein, "primer extension conditions" in the reaction mixture include conditions that render substantially all primer binding sites single-stranded. In some embodiments, such conditions are achieved by melting a double-stranded target nucleic acid, which renders the primer binding sites single-stranded, thereby enabling primers to anneal to these primer binding sites and form substrates for polymerase extension.

[0112] As exemplified in the references cited below, the predetermined positions or sites to which the first and second sets of primers bind can be determined by conventional methods known to those of ordinary skill in the art of multiplex nucleic acid amplification (e.g., multiplex PCR). For example, in the case of target polynucleotides being recombinant nucleic acids encoding immune receptor molecules, Faham and Willis (cited above), Van Dongen et al, Leukemia, 17:2257-2317 (2003) and similar references provide guidance for selecting primer binding sites for multiplex amplification of such target polynucleotides. In some embodiments, the selection of such predetermined positions or sites depends on several factors including: (i) the effect of such positions or sites on amplification efficiency (ideally, the frequency of different copies in the amplicon faithfully reflects the frequency of the target polynucleotide in the sample), (ii) the effect of such positions or sites on the copy length in the amplicon is consistent with the requirements of the DNA sequencing chemistry used, and (iii) whether the selected primers span a portion of the recombinant nucleic acid with the desired diversity, such as the VDJ region, etc. In this regard, the present invention relates to the understanding and recognition that because a set of sequences is a readout rather than an analog signal, cross-reactions of primers with different target polynucleotides do not affect the results of the methods of the present invention (compared to, for example, in methods based entirely on analog readouts, profiles, etc. of PCR amplification).

[0113] In some embodiments, the sequencing step includes the following steps: (i) providing a plurality of sequence reads, each sequence read having an error rate, and each sequence read comprising a nucleotide sequence and a tag sequence; and (ii) aligning a group of sequence reads having similar tag sequences, and then performing base calls within the group based on the sequence reads to determine the nucleotide sequence. Subsequently, such group-level nucleotide sequences may be merged into the same clonotype or different clonotypes as described below. In some embodiments, in the PCR step, the length of the receptor-specific portion of the first and second sets of primers is selected so that the relative content of different recombinant nucleic acids in the amplicon is substantially consistent with the relative content of the recombinant nucleic acids in the sample. When the primers are selected in this way, the primer binding sites on the respective target polynucleotides of the primers may have different positions and lengths. In some embodiments, the sequence tags are selected from a set of sequence tags that is much larger than the number of different target polynucleotides in the sample, such that substantially every different target polynucleotide and its copies in the sample will have a different sequence tag (e.g., in accordance with the "labeling by sampling" method described in Brenner, US patent 7,537,897). In some embodiments, the number of sequence tags in such a set is at least 100 times the size of the target polynucleotide population in the sample. In addition, in some embodiments where substantially every original target polynucleotide and its copies are labeled with the same unique sequence tag, the sequencing step includes: generating sequence reads of the nucleic acid of the amplicon, and aligning the sequence reads with the same sequence tag to determine the sequence reads corresponding to the same clonotype in the sample. In addition, in some embodiments, the aligning step further includes: determining the nucleotide sequence of each clonotype by determining the majority nucleotide at each nucleotide position of the sequence reads with the same sequence tag. In addition, in some embodiments, the step of removing the unextended primers can be performed by cleaving the single-stranded nucleic acid in the reaction mixture using a nuclease having 3'→5' single-stranded exonuclease activity (the nuclease can be provided by, for example, E. coli exonuclease I, which can be conveniently inactivated by heating). In other embodiments, the above method can be used to generate a clonotype profile to diagnose and / or monitor minimal residual lesions in cancer patients (such as myeloma, lymphoma or leukemia patients). After the above method steps, such diagnosis and / or monitoring can be performed using the following additional steps: determining the presence and / or level of one or more patient-specific clonotypes associated with the cancer from the clonotype profile. The method of this embodiment may also include the following steps: determining the sequence of each of the one or more sequence tags, and then comparing these sequences with the sequences of the sequence tags of the previously determined clonotype profile to determine the presence and / or level of contaminating sequences.In some embodiments, such comparing step comprises comparing the sequence of one of the more sequence tags to sequence tags of a clonotype database comprising clonotypes from at least one non-patient individual.

[0114] In yet another embodiment, a method for amplifying multiple recombinant nucleic acids encoding β, δ, and γ T cell receptor components in a single reaction may include the following steps: (a) mixing a first set of primers with a recombinant nucleic acid sample from a T cell under primer extension conditions in a reaction mixture, wherein each of the recombinant nucleic acids comprises at least a portion of a T cell receptor Jβ, Jδ, or Jγ segment at a first end, and wherein each primer of the first set has a receptor-specific portion having a length such that it anneals with a first end of a different recombinant nucleic acid and then extends to form a first extension product, and wherein each primer of the first set has a 5' non-complementary end comprising a sequence tag and a first primer binding site in a 3'→5' order, and the sequence tags are different (a) adding a second set of primers to the reaction mixture under primer extension conditions, each primer of the second set having a receptor-specific portion having a length that allows annealing with the first extension product and then extending to form second extension products, wherein each second extension product comprises at least a portion of a T cell receptor Vβ, Vδ or Vγ segment, and wherein each primer of the second set has a 5' non-complementary end comprising a second primer binding site; and (d) performing a polymerase chain reaction in the reaction mixture to form an amplicon, the polymerase chain reaction using a forward primer having specificity for the first primer binding site and a reverse primer having specificity for the second primer binding site. The above method may further include the step of sequencing a sample of the amplicon sequences. Such a sample is typically a "representative sample" because it is large enough that different clonal types are present in the sample at approximately the same frequency as they were present in the original sample of the biological material. In some embodiments, the sequencing step includes providing a plurality of sequence reads, each sequence read having an error rate, and each sequence read comprising a nucleotide sequence and a tag sequence, and then aligning sequence reads having similar tag sequences to determine sequence reads corresponding to the same clonotype. Whenever multiple sequence tags are attached to the original target polynucleotide or a copy thereof, such sequence reads may be processed in a further merging step, as described more fully below.

[0115] In another embodiment, a method for generating a clonal profile from a plurality of T cell receptor chains may include the following steps: (a) mixing a first set of primers with a recombinant nucleic acid sample from a T cell under primer extension conditions in a reaction mixture, wherein each primer of the first set has a receptor-specific portion having a certain length such that it anneals with a different recombinant nucleic acid at a predetermined position and then extends to form a first extension product, and wherein each primer of the first set has a 5' non-complementary end comprising a first primer binding site; (b) removing the unextended primers of the first set from the reaction mixture; (c) adding a second set of primers to the reaction mixture, wherein each primer of the second set has a receptor-specific portion having a certain length such that the receptor-specific portion is specific to the first extension product at a predetermined position and has a 5' non-complementary end comprising a second primer binding site, and the first set of primers and / or the second set of primers comprise a sequence tag located between the receptor-specific portion and the first or second primer binding site, respectively; and (d) performing a first polymerase chain reaction to form a first amplicon, wherein the first The polymerase chain reaction uses a forward primer specific to a first primer binding site and a second set of primers, wherein each nucleotide sequence of the first amplicon comprises the first primer binding site, the second primer binding site, at least one sequence tag, and comprises a portion of the Vβ segment and a portion of the Jβ segment of a T cell receptor chain, or a portion of the V segment and a portion of the Jδ segment of a T cell receptor chain, or a portion of the Vγ segment and a portion of the Jγ segment of a T cell receptor chain, and wherein the lengths of the receptor-specific portions of the first and second sets of primers are selected so that the relative amounts of different recombinant nucleic acids in the amplicon are substantially consistent with the relative amounts of different recombinant nucleic acids in the sample; (e) adding a reverse primer specific to the second primer binding site; (f) performing a second polymerase chain reaction in the reaction mixture to form a second amplicon, the polymerase chain reaction using a forward primer specific to the first primer binding site and a reverse primer specific to the second primer binding site; and (g) sequencing the nucleic acids of the second amplicon to form a clonal profile of multiple T cell receptor chains. In some embodiments, the sequencing step includes providing a plurality of sequence reads, each sequence read having an error rate, and each sequence read comprising a nucleotide sequence and a tag sequence, and then aligning sequence reads having similar tag sequences to determine sequence reads corresponding to the same clonotype. In other embodiments where the target polynucleotide and / or its copies are labeled with more than one sequence tag, the sequence reads may be processed in a further merging step as described more fully below.

[0116] As another example, a method for generating a clonal profile from a plurality of B cell receptor chains may be performed by the following steps: (a) mixing a first nested set of primers with a recombinant nucleic acid sample from a B cell under primer extension conditions in a reaction mixture, the first nested set comprising one or more primer groups, wherein each primer in each group has a receptor-specific portion, the receptor-specific portion having a length such that the receptor-specific portion of each primer from a different group anneals to a different recombinant nucleic acid at a predetermined site that does not overlap with a predetermined site of any other primer in the first nested set, and wherein each primer in each group has a 5' non-complementary end comprising a first primer binding site; (b) extending the primers of the first nested set to form a first extension product; (c) removing the unextended primers of the first nested set from the reaction mixture; and (d) adding a second nested set of primers to the reaction mixture under primer extension conditions, the second nested set comprising one or more primer groups, wherein each primer in each group has a receptor-specific portion, the receptor-specific portion having a length such that the receptor-specific portion of each primer from a different group anneals to a different recombinant nucleic acid. The method comprises the steps of: (a) annealing a first extension product to a first primer at a predetermined site that does not overlap with a predetermined site of any other primer of the second nested set, wherein each primer of each group has a 5' non-complementary end comprising a second primer binding site, and wherein the first nested set of primers and / or the second nested set of primers comprise a sequence tag located between their receptor-specific portion and their first or second primer binding site, respectively; (e) extending the second nested set of primers to form second extension products, such that each second extension product comprises the first primer binding site, the second primer binding site, at least one sequence tag, and any of the following: (i) a portion of the V segment and a portion of the J segment of a B cell receptor heavy chain, or (ii) a portion of the V segment and a portion of the J segment of a B cell receptor kappa light chain; (f) performing a polymerase chain reaction in a reaction mixture to form an amplicon, the polymerase chain reaction using a forward primer specific for the first primer binding site and a reverse primer specific for the second primer binding site; and (g) sequencing the nucleic acid of the amplicon to form a clonal profile of multiple B cell receptor chains.

[0117] In some embodiments, the primers may be subjected to more than one cycle of annealing and extension (after melting the extension products) in steps (b) and / or (e), in which case the copies of the original recombinant nucleic acid in the sample may be labeled with one or more sequence tags. In these embodiments, the sequencing step (g) may include further alignment and merging steps as described below for determining clonotypes and clonotype profiles. In some embodiments, for example, in embodiments where only a single extension is performed in steps (b) and (e), the sequencing step includes: providing a plurality of sequence reads, each sequence read having an error rate, and each sequence read comprising a nucleotide sequence and a tag sequence, and then aligning sequence reads having similar tag sequences to determine sequence reads corresponding to the same clonotype. As above, in some embodiments, the position and length of the receptor-specific portion of the first and second sets of primers are selected in PCR so that the relative amounts of different recombinant nucleic acids in the amplicon are substantially consistent with the relative amounts of different recombinant nucleic acids in the sample.

[0118] In some embodiments, sequence tags are attached to target polynucleotides or copies thereof in a primer extension step, wherein substantially each different target polynucleotide and copy thereof is labeled with the same sequence tag. In other embodiments, target polynucleotides or copies thereof of a sample may be labeled with more than one different sequence tag. As will be further described below, in some embodiments, multiple extensions or multiple PCR cycles may be performed in the presence of primers (first set of primers or second set of primers) comprising sequence tags, which may cause different sequence tags to be attached to the same target polynucleotide and / or a copy thereof.

[0119] Sequence tags in clonotype analysis

[0120] In one aspect, the present invention relates to a method for obtaining and analyzing sequence data from an immune molecule repertoire such as a T cell receptor (TCR) or a B cell receptor (BCR) or a determined fragment thereof to quickly and efficiently determine a clonal type spectrum. The sequence data typically includes a large number of sequence reads obtained from a DNA sequencer for analyzing immune molecules, i.e., sequences of base calls and related quality scores. The key challenge in constructing a clonal type spectrum is that some sequence reads contain errors from non-biological sources (such as extraction steps, sequencing chemistry, amplification chemistry, etc.), and sequence reads with true differences need to be quickly and accurately identified from these sequence reads. One aspect of the present invention includes attaching a unique sequence tag to each target polynucleotide (e.g., a recombinant nucleic acid) in a sample to help determine whether the sequence reads of such conjugates are derived from the same original target polynucleotide. According to one aspect of the present invention, a sequence tag is attached to a somatic cell recombinant nucleic acid molecule to form a tag-molecule conjugate, wherein each recombinant nucleic acid of such a conjugate has a unique sequence tag. This attachment is usually performed after the nucleic acid molecule is extracted from a sample containing T cells and / or B cells and / or cell-free DNA. Preferably, such unique sequence tags are as different from each other as possible, as determined by conventional sequence distance measures (e.g., Hamming distance, etc.). By maximizing the distance between sequence tags in a tag-molecule conjugate, a sequence tag of one conjugate is always much closer to its ancestral tag sequence than to the ancestral tag sequence of any other tag sequence in a different conjugate, even if sequencing and amplification error rates are high. For example, if 16-mer sequence tags are used, and the Hamming distance between each such tag on a set of clonotypes and all other sequence tags on the set of clonotypes is at least 50% (i.e., eight nucleotides), then at least eight sequencing or amplification errors will be required to convert one such tag to another, thereby causing a misreading of the sequence tag (and incorrect grouping of sequence reads of the clonotypes according to the erroneous sequence tag). In one embodiment, the sequence tags are selected so that after the tags are attached to the recombinant nucleic acid molecules to form tag-molecule conjugates, the value of the Hamming distance between the tags on the tag-molecule conjugates is at least 25% of the total length of such sequence tags (that is, each sequence tag differs from all other such tags in sequence by at least 25% of the nucleotides); in another embodiment, the value of the Hamming distance between such sequence tags is at least 50% of the total length of such sequence tags.

[0121] In one aspect, the present invention is implemented by the following steps: (a) obtaining a sample containing T cells and / or B cells and / or cell-free DNA from an individual; (b) attaching a sequence tag to a recombinant nucleic acid molecule of a T cell receptor gene or an immunoglobulin gene in the sample to form a tag-molecule conjugate, wherein substantially each molecule of the tag-molecule conjugate has a unique sequence tag; (c) amplifying the tag-molecule conjugate; (d) sequencing the tag-molecule conjugate; and (e) aligning sequence reads of similar sequence tags to determine sequence reads corresponding to the same recombinant nucleic acid in the sample. A sample containing B cells or T cells is obtained using conventional techniques. In the step of attaching sequence tags, it is preferred that the sequence tags are not only unique, but also completely different from each other, and even if a large number of sequencing or amplification errors occur, the possibility of converting one sequence tag into another will be close to zero. For most sequencing technologies, the tag-molecule conjugate must be amplified after attaching the sequence tag; however, whenever a single molecule sequencing technology is used, the amplification step is optional. Single-molecule sequencing technology includes but is not limited to single-molecule real-time (SMRT) sequencing, nanopore sequencing, etc., for example, U.S. Patents 7,313,308, 8,153,375, 7,907,800, 7,960,116, 8,137,569; Manrao et al, Nature Biotechnology, 4(8): 2685-2693 (2012) (Manrao et al., Nature Biotechnology, 2012, Vol. 4, No. 8, pp. 2685-2693), etc.

[0122] On the other hand, the present invention includes a method for determining the number of lymphocytes in a sample by counting unique sequence tags. Even without sequence tags, clonotypes of TCRβ or IgH genes, especially those containing V(D)J regions, provide unique markers for lymphocytes and their clones. Whenever a recombinant nucleic acid is obtained from genomic DNA, the unique clonotype can be counted after sequencing, and the number of clonotypes obtained is used to estimate the lymphocyte count in the sample. Whenever there is a significant clonal population of the same lymphocyte associated with the same clonotype (or when the recombinant nucleic acid is obtained from a sample mRNA whose amount of individual sequences can reflect or depend on expression rate and cell number), the method fails. The use of sequence tags overcomes this defect and is particularly suitable for lymphocyte counting in patients with a variety of lymphatic diseases (such as lymphoma or leukemia). According to one aspect of the present invention, sequence tags can be used to obtain an absolute count of lymphocytes in a sample regardless of whether there is a large dominant clone, such as with leukemia. Such methods can be implemented by the following steps: (a) obtaining a sample containing lymphocytes from an individual; (b) attaching a sequence tag to a recombinant nucleic acid molecule of a T cell receptor gene or an immunoglobulin gene of a lymphocyte to form a tag-molecule conjugate, wherein substantially each molecule of the tag-molecule conjugate has a unique sequence tag; (c) amplifying the tag-molecule conjugate; (d) sequencing the tag-molecule conjugate; and (e) counting the number of different sequence tags to determine the number of lymphocytes in the sample. In some embodiments, the recombinant nucleic acid molecule is from genomic DNA.

[0123] In one embodiment of the present invention, sequence tags are attached to recombinant nucleic acid molecules of a sample by sampling labeling, such as disclosed in Brenner et al, US patent 5,846,719 (Brenner et al., U.S. Patent 5,846,719); Brenner et al, US patent 7,537,897 (Brenner et al., U.S. Patent 7,537,897); Macevicz, International patent publication WO 2005 / 111242 (Macevicz, International Patent Publication WO 2005 / 111242), etc., which are incorporated herein by reference. In sampling labeling, a population of polynucleotides to be labeled (or uniquely labeled) is used to sample (by attachment, connection, etc.) sequence tags of a much larger population. That is, if the polynucleotide population has K members (including repeats of the same polynucleotide) and the population of sequence tags has N members, then N>>K. In one embodiment, the population size of sequence tags used by the present invention is at least 10 times the size of the population of clonotypes in the sample; in another embodiment, the population size of sequence tags used by the present invention is at least 100 times the size of the population of clonotypes in the sample; and in another embodiment, the population size of sequence tags used by the present invention is at least 1000 times the size of the population of clonotypes in the sample. In other embodiments, the population size of sequence tags is selected so that whenever such clonotype is combined with such a population of sequence tags in (for example) an attachment reaction (such as a ligation reaction), an amplification reaction, etc., substantially each clonotype in the sample will have a unique sequence tag. In some embodiments, substantially each clonotype means that at least 90% of such clonotypes will have a unique sequence tag; in other embodiments, substantially each clonotype means that at least 99% of such clonotypes will have a unique sequence tag; in other embodiments, substantially each clonotype means that at least 99.9% of such clonotypes will have a unique sequence tag. In many tissue samples or biopsies, the number of T cells or B cells can be as high as or about 1 million cells; therefore, in some embodiments of the present invention using such samples, the number of unique sequence tags used in the sampling labeling is at least 10 8 , or in other embodiments at least 10 9 .

[0124] In such embodiments where up to 1 million clonotypes are sampled for labeling, combinatorial synthesis can be performed by reacting a mixture of all four nucleotide precursors in each addition step of the synthesis reaction, thereby efficiently generating a large set of sequence tags, such as disclosed in Church, US patent 5,149,625, which is incorporated by reference. The result is a sequence tag having "N1N2...N k " structured sequence tag group, where each N i =A, C, G or T, and k is the number of nucleotides in the tag. The number of sequence tags in the sequence tag set generated by this combination synthesis is 4 k Thus, such a set of sequence tags with k being at least 14 or k being in the range of about 14 to 18 is suitable for attaching sequence tags to 10 6 On the molecular population of each member. The sequence tag group with the above structure includes many sequences, and the sequence may cause difficulties or errors when implementing the inventive method. For example, the sequence tag group synthesized by the above combination includes many member tags with homopolymer segments, so some sequencing methods (such as synthetic sequencing methods) are difficult to accurately measure when the homopolymer segment exceeds a certain length. Therefore, the present invention includes a sequence tag synthesized by combination, which has a structure effective for a specific method step (such as sequencing). For example, four kinds of natural nucleotides can be divided into non-intersecting subsets, and these non-intersecting subsets are used alternately in the combination synthesis, thereby producing several sequence tag structures effective for synthetic sequencing chemical reactions, thereby preventing the homopolymer segment from exceeding a given length. For example, z is made A or C, and x is made G or T, to produce a sequence tag with the following structure:

[0125] [(z)1(z)2...(z) i ][(x)1(x)2...(x) j ]...

[0126] Wherein, i and j may be the same or different, and are selected to limit the size of any homopolymer segment. In one embodiment, i and j are in the range of 1 to 6. In such embodiments, the length of the sequence tag may be in the range of 12 to 36 nucleotides; and in other embodiments, the length of such sequence tags may be in the range of 12 to 24 nucleotides. In other embodiments, other nucleotide pairings may be used, for example, z is A or T, and x is G or C; or z is A or G, and x is T or C. Alternatively, z'' is a combination of any three of the four natural nucleotides, and x' is any nucleotide that is not z' (for example, z' is A, C or G, and x' is T). This produces the following sequence tag structure:

[0127] [(z')1(z')2...(z')i ]x'[(z')1(z')2...(z') i ]x'...

[0128] wherein i is selected as described above and the presence of x' serves as a punctuation to terminate any undesired homopolymers.

[0129] Other sequence tags

[0130] The present invention uses a method of labeling nucleic acids (such as fragments of genomic DNA) with unique sequence tags, followed by amplification and sequencing, and the tags may include "mosaic tags". Such sequence tags can be used to identify amplification and sequencing errors. Mosaic tags minimize the artifacts of sequencing and amplification, which are caused by improper annealing, priming, hairpin structure formation, etc., and may appear with completely random sequence tags of the prior art. In one aspect, the mosaic tag is a sequence tag comprising alternating constant regions and variable regions, wherein each constant region is located in the mosaic tag and comprises a predetermined nucleotide sequence, and each variable region is located in the mosaic tag and comprises a predetermined number of randomly selected nucleotides. By way of example, a 22-mer mosaic tag (SEQ ID NO: 1) may have the following form:

[0131] Nucleotide position:

[0132]

[0133] Regional Location

[0134] There are nine constant and variable regions, of which region 1 (nucleotides 1-3), region 3 (nucleotides 9), region 5 (nucleotides 12-14), region 7 (nucleotides 18-19) and region 9 (nucleotides 21-22) are variable regions (nucleotides indicated by double underlines), and region 2 (nucleotides 4-8), region 4 (nucleotides 10-11), region 6 (nucleotides 15-17) and region 8 (nucleotide 20) are constant regions. N represents a nucleotide randomly selected from A, C, G or T; therefore, the number of mosaic tags in this example is 4 11 = 4,194,304 tags, b represents the predetermined nucleotide at the indicated position. In some embodiments, the sequence of b "***bbbb*bb***bbb**b**" is selected to minimize the possibility of a perfect match in the genome of the organism constituting the sample.

[0135] In one aspect, for the mosaic tags in a specific embodiment of the method of the invention, all constant regions in the same position have the same length, and all variable regions in the same position have the same length. This allows the mosaic tags to be synthesized by partial combinatorial synthesis using conventional chemical reactions and instruments.

[0136] In one aspect, the mosaic tag comprises 10 to 100 nucleotides, or 12 to 80 nucleotides, or 15 to 60 nucleotides. In some embodiments, the mosaic tag comprises at least eight nucleotide positions with randomly selected nucleotides; in other embodiments, whenever the mosaic tag is at least 15 nucleotides in length, it comprises at least 12 nucleotide positions with randomly selected nucleotides. In another aspect, the length of the variable region within the mosaic tag may not exceed seven nucleotides.

[0137] On the other hand, mosaic tags can be used in the following steps: (i) preparing DNA templates from nucleic acids in a sample; (ii) sampling and labeling DNA templates to form multiple tag-template conjugates, wherein substantially each DNA template of the tag-template conjugate has a unique mosaic tag, the mosaic tag comprising alternating constant regions and variable regions, each constant region being located in the mosaic tag and comprising a predetermined sequence of 1 to 10 nucleotides in length, and each variable region being located in the mosaic tag and having a length of 1 to 10 randomly selected nucleotides, such that constant regions at the same position have the same length, and variable regions at the same position have the same length; (iii) amplifying the multiple tag-template conjugates; (iv) generating multiple sequence reads for each amplified tag-template conjugate; and (v) determining the common nucleotides at each nucleotide position of each group of multiple sequence reads having the same mosaic tag, thereby determining the nucleotide sequence of each nucleic acid. In another aspect, a mosaic tag can be used in the following steps: (a) preparing a single-stranded DNA template from a nucleic acid in a sample; (b) sampling and labeling the single-stranded DNA template to form a tag-template conjugate, wherein substantially each single-stranded DNA template of the tag-template conjugate has a unique sequence tag (i.e., a mosaic tag), and the unique sequence tag has a length of at least 15 nucleotides and has the following form:

[0138] [(N1N2...N Kj )(b1b2...b Lj )] M

[0139] Among them, each N i (i=1,2,...K j ) is a nucleotide randomly selected from A, C, G and T; when each j is less than or equal to M (i.e., the region N1N2...N Kj When is the variable region, K jis an integer in the range of 1 to 10; each b i (i=1,2,...L j ) is a nucleotide; when each j is less than or equal to M, L j is an integer in the range of 1 to 10; thus, each sequence tag (i) has the same K for each j. j , and (ii) for each j has the same sequence b1b2...b Lj (i.e. area b1b2...b Lj is a constant region); and M is an integer greater than or equal to 2; (c) amplifying the tag-template conjugate; (d) generating a plurality of sequence reads for each amplified tag-template conjugate; and (e) determining the common nucleotides at each nucleotide position of each group of multiple sequence reads having the same sequence tag, thereby determining the nucleotide sequence of each nucleic acid. In some embodiments, the plurality of sequence reads is at least 10 4 In other embodiments, the plurality of sequence reads is at least 10 5 In other embodiments, the plurality of sequence reads is at least 10 6 In some embodiments, the total length of the sequence tags is in the range of 15 to 80 nucleotides.

[0140] Attach sequence tags

[0141] A variety of different attachment reactions can be used to attach a unique tag to substantially every clonotype in a sample (in addition to those shown above). Many techniques for capturing subsets of sample nucleic acids to, for example, reduce sample complexity in microarray or genomic sequencing technologies can be used with conventional modifications of the present invention for attaching sequence tags to recombinant nucleic acids. Exemplary techniques for capturing different sets of target nucleic acids for subsequent manipulation (including attachment of sequence tags, sequencing, etc.) include the following: Willis et al, US patent 7,700,323; Jones et al, US patent publication 2005 / 0142577; Gullberg et al, US patent publication 2005 / 0037356; Porreca et al, Nature Methods, 4(11):931-936 (2007); Turner et al, Nature Methods, 4(11):931-936 (2007); al, Nature Methods, 6(5): 315-316 (2009) (Turner et al., Nature Methods, 2009, Vol. 6, No. 5, pp. 315-316); Church, US patent 5,149,625 (Church, US patent 5,149,625); Macevicz, US patent 8,137,936 (Macevicz, US patent 8,137,936), etc.

[0142] In one embodiment, a sample comprising a recombinant nucleic acid molecule (which in turn comprises a clonal sequence) is combined with a population or library of sequence tags so that members of the two molecular populations can be randomly combined and (e.g., covalently) associated or connected, thereby completing such attachment. For example, such random combinations may occur in a bimolecular reaction, wherein a primer comprising a tag is annealed to a target nucleic acid and extended, or an adapter comprising a tag is connected to the end of the target nucleic acid. In some embodiments, the method of attaching a tag may depend in part on the method of DNA sequencing. For example, in a sequencing method (such as 454 sequencing) that produces relatively long accurate sequence reads, a cDNA library can be made from mRNA comprising recombinant nucleic acids using conventional techniques (such as 5'-RACE), as disclosed in Freeman et al, Genome Research, 19: 1817-1824 (2009) (Freeman et al., Genome Research, 2009, Vol. 19, pp. 1817-1823), and then an adapter comprising a sequence tag can be connected to one or both ends, thereby attaching a sequence tag. In other embodiments, when sequencing methods (such as Illumina sequencing or Ion Torrent sequencing) are used to generate relatively short and error-prone sequence reads, further measures may need to be taken so that the amplicons used for sequencing have the length covered by the sequence reads generated by the technology. In such tag attachment reactions, the clonotype sequence includes linear single-stranded or double-stranded polynucleotides, and the sequence tags are carried by reagents such as amplification primers (such as PCR primers), ligation adapters, circularizable probes, plasmids, etc. Several such reagents capable of carrying a population of sequence tags are disclosed in the following documents: Macevicz, US patent 8,137,936; Faham et al, US patent 7,862,999; Landegren et al, US patent 8,053,188; Unrau and Deugau, Gene, 145: 163-169 (1994); Church, US patent 5,149,625, etc., which are incorporated herein by reference.

[0143] Figure 2A and 2B An attachment reaction involving PCR is shown, in which a population of sequence tags (Τ1, T2, T3 ... T j,T j+i ...T k ,T k+l ...T n-1 ,T n ) is incorporated into primer 2100. The size of the sequence tag population is much larger than the size of the recombinant nucleic acid molecule 2102. The sequence tag is attached to the recombinant nucleic acid molecule by annealing the primer to the nucleic acid molecule in the first cycle of PCR and extending the primer with a DNA polymerase. The figure depicts the random annealing of the recombinant nucleic acid molecule with the primer via its universal primer binding region 2104 in (for example) V region 2108, thereby selecting (or sampling) a small part of the entire sequence tag population. Since the primer (and thus the sequence tag) is randomly combined with the recombinant nucleic acid sequence molecule, the possibility of the same sequence tag being attached to different nucleic acid molecules is small; however, if the sequence tag population is as large as described herein, then this possibility will be extremely small, so that essentially each recombinant nucleic acid molecule will be attached with a unique sequence tag. The other primer 2106 of the forward and reverse primer pair anneals to the C region 2110 so that an amplicon 2112 is formed after multiple cycles of annealing, extension, and melting, thereby attaching a unique sequence tag to the V(D)J region that includes the clonotypes of the population 2102. That is, the amplicon 2112 includes the tag-molecule conjugate from the attachment reaction.

[0144] Figure 2C and 2D A method for attaching a pair of sequence tags to each or substantially each recombinant nucleic acid in a sample is shown. Figure 2A and 2B In the method shown, the sequence tags (Τ1, T2, T3...T j ,T j+i ...T k ,T k+ 1...T n-1 ,T n ) was used as the downstream primer. In addition, a primer 2200 with a sequence tag (T m ,T m+1 ,T m+2 ...T q ,T q+1 ,T q+2 ,...T r ,T r+1 ,T r+2 ,...T s ,T s+1 ,T s+2, ...) is used as the upstream primer instead of the universal primer 2106. As with the downstream primer set, the number of different sequence tags carried by the upstream primer 2206 can be large compared to the number of recombinant nucleic acid molecules 2202, so that after amplification, substantially each recombinant nucleic acid 2202 will have a unique tag. In some embodiments, each set of sequence tags in primers 2206 and 2200 need not be identical. Figure 2A and 2B The sequence tag groups in the illustrated embodiment are the same size. Since each recombinant nucleic acid is uniquely labeled by a pair of sequence tags, for a pair of sequence tags that label a single recombinant nucleic acid, sharing one sequence tag in the pair of sequence tags with a different recombinant nucleic acid does not detract from its substantial uniqueness. Figure 2C and 2D In an embodiment, the sequence tags of each primer set 2200 and 2206 may be less diverse than the sequence tags of primer set 2100. For example, if random sequence tags are used, and primer 2100 includes a 16-mer sequence tag, then primers 2200 and 2206 may each include an 8-mer sequence tag to obtain the same overall sequence tag diversity. Figure 2C and 2D Examples and Figure 2A and 2B The embodiment of operates in a similar manner. The sequence tag is attached to the recombinant nucleic acid molecule by annealing the primer to the nucleic acid molecule in the first cycle of PCR and extending the primer with a DNA polymerase. As above, Figure 2C Describes the manner in which a recombinant nucleic acid molecule is randomly annealed and paired with a primer via its universal primer binding region 2204 and 2205 in, for example, V region 2208 and C region 2210, thereby selecting (or sampling) a small portion of the entire sequence tag pair population. Due to the random combination of primers (and thus sequence tags) with recombinant nucleic acid sequence molecules, the probability of the same sequence tag pair being attached to different nucleic acid molecules is small; however, if the sequence tag population is as large as described herein, then this probability will be extremely small, so that essentially each recombinant nucleic acid molecule will be attached with a unique sequence tag pair. Amplicon 2212 is formed after multiple annealing, extension and melting cycles, thereby attaching a unique sequence tag pair to a V(D)J region, which includes a clonal type of population 2202.

[0145] That is, the amplicon 2212 includes the tag-molecule conjugate from the attachment reaction. In some embodiments, a circularizable probe can be used to capture the sequence tag and attach the sequence tag to the desired recombinant nucleic acid by conventional modifications of the techniques disclosed, for example, by Porreca et al. (cited above), Willis et al. (cited above), or similar references. Figure 2E and2F As shown, a circularizable probe 2302 is provided that comprises the following elements: an upstream target binding segment 2304, a downstream target binding segment 2306 having a 5'-phosphorylated end 2305; a sequence tag 2310; a second universal primer binding site 2314; an optional cleavage site 2308; and a first universal primer binding site 2312. Under annealing conditions, the circularizable probe 2302 is combined with a sample comprising a target polynucleotide 2300 in a reaction mixture, the target polynucleotide being, for example, the first strand or the second strand of a cDNA prepared from mRNA by conventional techniques. As shown, the target polynucleotide comprises a V region, a NDN region, a J region, and a C region of a recombinant nucleic acid encoding an IgH or TCR β chain. In some embodiments, the sequences of the upstream target binding segment 2304 and the downstream target binding segment 2306 are selected so that they span a portion of the VDJ region of the target polynucleotide. After annealing the upstream target binding segment 2304 and the downstream target binding segment 2306, the circularizable probe 2302 and the target polynucleotide 2300 form a complex 2330 in the reaction mixture. In the presence of a DNA polymerase and dNTPs, the upstream target binding segment 2304 extends (2340) to the downstream target binding segment 2306, replicating (thereby capturing) a portion of the VDJ region of the target polynucleotide. When there is ligase activity, the extended upstream target binding segment is connected to the downstream target binding segment 2306, thereby forming a closed single-stranded DNA loop 2342. The reaction mixture can then be optionally treated (2344) with an exonuclease to remove unreacted probes and target polynucleotides. In some embodiments, the single-stranded loop 2342 is linearized by cutting at the cutting site 2308 (the cutting site can be, for example, a rare-cutting endonuclease recognition site), or inserting an RNA monomer into the probe and cutting with RNase H or a similar enzyme, and then the VDJ tag insertion sequence of the linearized probe 2348 can be amplified by primers 2350 and 2352. Primers 2350 and 2352 can include non-complementary regions that are used to add elements so that DNA sequencing can be performed later (2354). Alternatively, the single-stranded loop can be used to generate a nanosphere template for direct sequencing, such as Drmanac et al, Science, 327 (5961): 78-81 (2010); U.S. Patent 8,445,196, etc.

[0146] Figure 2GAnother embodiment of attaching a sequence tag to a recombinant nucleic acid encoding an immune receptor molecule is shown. Guidance for implementing this embodiment can be found in Faham and Zheng, US patent 7,208,295, which is incorporated herein by reference. Under annealing conditions, the recombinant nucleic acid 2450 is combined with a probe 2454 and an adapter 2456 in a reaction mixture. The probe 2454 comprises a receptor-specific portion 2455 and an adapter-specific portion 2457. For example, the probe 2454 may include a probe mixture in which different probes have receptor-specific portions that are specific to different J regions, or in other embodiments, the receptor-specific portions are specific to different V regions. The 5'-phosphorylated adapter 2456 comprises a probe-specific portion 2458, a sequence tag 2460, and a first primer binding site 2462 located at its 5' end. The position, sequence and length of the receptor-specific portion 2455 and the adapter-specific portion 2457 and the probe-specific portion 2458 of the probe 2454 are selected so that they hybridize to each other to form the structure 2452. After the structure 2452 is formed, the single-stranded portion 2461 is cut from the recombinant nucleic acid 2450 and the free 3' end of the recombinant nucleic acid 2450 is ligated to the 5' phosphorylated end of the adapter 2456 to form a first extension product 2459, followed by removal (2474) of the probe 2454. Cleavage of 2461 can be accomplished using a single-stranded nuclease as described in the Faham and Zheng patents. In one embodiment, probe 2454 is synthesized by replacing thymidine with uracil in a PCR, for example, by substituting dUTP for dTTP, and removing uracil by treatment with uracil DNA glycosylase (UDG), such as described in Faham et al, US patent 7,208,295, which is incorporated by reference. UDG treatment cleaves uracil in probe 2454 to produce fragment 2455. After releasing the probe, the adapter and wings are removed (2476), forward primer 2466 and reverse primer 2468 are added to the extension product 2464, and PCR is performed (2470), and a sample of the resulting amplicons is then sequenced (2472).

[0147] In with Figure 2GIn an embodiment similar to the illustrated embodiment, a similar probe and adapter can be used to attach a sequence tag to a predetermined site of a target polynucleotide, wherein a flap endonuclease (e.g., FEN-1) is used to cleave the single-stranded portion corresponding to 2461. In this embodiment, in addition to using a different nuclease, the polarity of the probe and adapter sequences is reversed; that is, the substrate for the flap endonuclease requires that the 3' end of the adapter corresponding to 2454 anneals to the target sequence 2450, and that the single-stranded portion corresponding to 2452 is the 5' end of the target sequence. After cleavage and removal of the probe sequence, the remaining steps are essentially the same. Guidance for the use of flap endonucleases in detection assays can be found in the following references: Lyamichev et al, Nature Biotechnology, 17:292-296 (1999); Eis et al, Nature Biotechnology, 19:673-676 (2001); and similar references.

[0148] In some embodiments, the recombinant nucleic acid encodes chains of immune receptor molecules that typically form an immune repertoire that may include a very large set of polynucleotides (e.g., >1000, and more commonly greater than 10,000, and more commonly 100,000 to 1,000,000, or more) that are very similar and may be less than 500 nucleotides in length, or less than 400 nucleotides in other embodiments, or less than 300 nucleotides in other embodiments. In one aspect of the invention, the inventors recognize and appreciate that these properties enable completely different sequence tags to be used to effectively compare sequence reads of highly similar clonotypes to determine whether they are derived from the same original sequence.

[0149] sample

[0150] The term "sample" refers to a quantity of biological material, which in some embodiments is derived from a patient and contains cells and / or cell-free DNA; that is, the term is used interchangeably with the term "specimen" or "tissue sample". The term "sample" is also sometimes used in the statistical sense of obtaining a subset or portion of a larger group or amount of, for example, recombinant nucleic acids, respectively; specifically, the statistical use of the term "sample" can also be understood to mean a "representative sample", in which case such a sample is understood to reflect or approximate the relative frequencies of different nucleic acids in, for example, a tissue. One skilled in the art will be able to discern the correct use of the term from the context.

[0151] A clonal type spectrum can be obtained from a sample of immune cells or fluid (such as blood), the sample comprising a cell-free nucleic acid encoding an immune receptor chain. For example, immune cells may include T cells and / or B cells. T cells (T lymphocytes) include, for example, cells expressing T cell receptors. T cells include helper T cells (effector T cells or Th cells), cytotoxic T cells (CTLs), memory T cells, and regulatory T cells. In one aspect, the T cell sample comprises at least 1,000 T cells; more typically, the sample comprises at least 10,000 T cells, and more typically comprises at least 100,000 T cells. On the other hand, the number of T cells contained in the sample ranges from 1000 to 1,000,000 cells. The sample of immune cells may also include B cells. B cells include, for example, plasma B cells, memory B cells, B1 cells, B2 cells, marginal zone B cells, and follicular B cells. B cells may express immunoglobulins (antibodies, B cell receptors). As above, in one aspect, the B cell sample comprises at least 1,000 B cells; more typically, the sample comprises at least 10,000 B cells, and more typically comprises at least 100,000 B cells. In another aspect, the sample comprises a number of B cells ranging from 1000 to 1,000,000 B cells.

[0152] The sample used in the method of the present invention can be from a variety of tissues, including (for example) tumor tissue, blood and plasma, lymph, cerebrospinal fluid around the brain and spinal cord, synovial fluid around bone joints, etc. In one embodiment, the sample is a blood sample. The blood sample can be about 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5 or 5.0 mL. The sample can be a tumor biopsy. The biopsy can be from (for example) a tumor of the brain, liver, lung, heart, colon, kidney or bone marrow. Any biopsy technique used by those skilled in the art can be used to separate and obtain a sample from a subject. For example, a biopsy can be an open biopsy, in which general anesthesia is used. The biopsy can be a closed biopsy, in which the incision caused is smaller than in an open biopsy. The biopsy can be a core biopsy or an incisional biopsy, in which a part of the tissue is removed. The biopsy may be an excisional biopsy, in which an attempt is made to remove the entire lesion. The biopsy may be a fine needle aspiration biopsy, in which a needle is used to remove a sample of tissue or fluid.

[0153] In some embodiments, for diagnostic samples, the clonotype spectrum of the method of the present invention is derived from a tumor or peripheral blood, or for samples used to monitor residual lesions, the clonotype spectrum is derived from peripheral blood. One or more clonotypes associated with a disease (e.g., a lymphatic disease or a myeloproliferative disease) are determined by a diagnostic sample. Typically, one or more clonotypes associated with a lymphatic disease or a myeloproliferative disease are those with the highest frequency of occurrence in the clonotype spectrum. In some cases, there may be a single associated clonotype, while in other cases, there may be multiple clonotypes associated with a lymphatic disease or a myeloproliferative disease. Tumor samples can be obtained from any tissue infected by such diseases, including lymph nodes or other tissues outside the lymphatic system. As described above, the clonotype spectrum for monitoring residual lesions can be derived from a nucleic acid sample extracted from peripheral blood. The nucleic acid of the sample can be from a B cell, and the B cell is from a cell-containing portion of peripheral blood (e.g., plasma), or from a cell-free portion of peripheral blood (e.g., serum). In one embodiment, the peripheral blood sample contains at least 1,000 B cells; more typically, the sample contains at least 10,000 B cells, and more typically contains at least 100,000 B cells. On the other hand, the number of B cells contained in the sample is in the range of 1000 to 1,000,000 B cells. In some embodiments, the number of cells in the sample limits the sensitivity of the measurement. That is, using a larger peripheral blood sample will make the sensitivity of detecting residual lesions higher. For example, in a sample containing 1,000 B cells, no matter how many sequencing reads are obtained when sequencing the DNA of these cells, the minimum frequency of detectable clonal types is 1 / 1000 or 0.001. The nucleic acid of the sample can come from T cells, which are from a cell-containing portion of peripheral blood (such as plasma), or from a cell-free portion of peripheral blood (such as serum). In one embodiment, the peripheral blood sample contains at least 1,000 T cells; more typically, the sample contains at least 10,000 T cells, and more typically contains at least 100,000 T cells. On the other hand, the number of T cells contained in the sample is in the range of 1000 to 1,000,000 T cells. In some embodiments, the number of cells in the sample limits the sensitivity of the measurement. That is, using a larger peripheral blood sample will result in a higher sensitivity for detecting residual lesions. For example, in a sample containing 1,000 T cells, no matter how many sequencing reads are obtained when sequencing the DNA of these cells, the minimum frequency of detectable clonal types is 1 / 1000 or 0.001.

[0154] The sample used in the present invention may include DNA (such as genomic DNA) or RNA (such as messenger RNA). The nucleic acid may be cell-free DNA or RNA (for example, extracted from the circulatory system), Vlassov et al, Curr. Mol. Med., 10: 142-165 (2010) (Vlassov et al, Current Molecular Medicine, 2010, Vol. 10, pp. 142-165); Swarup et al, FEBS Lett., 581: 795-799 (2007) (Swarup et al, FEBS Lett., 581: 795-799 (2007) (Swarup et al, FEBS Newsletter, 2007, Vol. 581, pp. 795-799). In the provided methods of the present invention, the amount of RNA or DNA available for analysis from a subject includes, for example, as low as a single cell in some applications (e.g., calibration tests using other cell selection criteria such as morphological indicators), and as high as 10 million cells or more, the 10 million or more cells are converted to DNA in the range of 6 pg to 60 μg, and the amount converted to RNA is in the range of 1 pg to 10 μg. In some embodiments, the nucleic acid sample is a DNA sample of 6 pg to 60 μg. In other embodiments, the nucleic acid sample is a DNA sample from 100 μL to 10 mL of peripheral blood; in other embodiments, the nucleic acid sample is a DNA sample from a cell-free portion of 100 μL to 10 mL of peripheral blood.

[0155] In some embodiments, the sample of lymphocytes or cell-free nucleic acid is large enough so that substantially every B cell or T cell with a distinct clonotype is represented therein, thereby forming a clonotype "repertoire". In one embodiment, to obtain a basic representation of each distinct clonotype, the sample taken contains every clonotype in the population with a frequency of 0.001% or greater with a probability of 99%. In another embodiment, the sample taken contains every clonotype in the population with a frequency of 0.0001% or greater with a probability of 99%. And in another embodiment, the sample taken contains every clonotype in the population with a frequency of 0.00001% or greater with a probability of 99%. In one embodiment, the sample of B cells or T cells contains at least 500,000 cells, and in another embodiment, such a sample contains at least 1 million cells.

[0156] Nucleic acid samples can be obtained from peripheral blood using conventional techniques, such as Innis et al, editors, PCR Protocols (Academic Press, 1990). For example, leukocytes can be isolated from a blood sample using conventional techniques, such as the Rosette Sep kit (Stem Cell Technologies, Vancouver, Canada). The volume of the blood sample can range from 100 μL to 10 mL; in one aspect, the blood sample volume ranges from 100 μL to 2 mL. DNA and / or RNA can then be extracted from such blood samples using conventional techniques suitable for the methods of the present invention, such as the DNeasy Blood & Tissue Kit (Qiagen, Valencia, CA). Optionally, subsets of leukocytes (e.g., lymphocytes) can be further separated using conventional techniques such as fluorescence activated cell sorting (FACS) (Becton Dickinson, San Jose, CA), magnetic activated cell sorting (MACS) (Miltenyi Biotec, Auburn, CA), etc. For example, memory B cells can be isolated via surface markers CD19 and CD27.

[0157] Cell-free DNA can also be extracted from peripheral blood samples using conventional techniques such as Lo et al, US patent 6,258,540 (Lo et al., U.S. Patent 6,258,540); Huang et al, Methods Mol. Biol., 444: 203-208 (2008) (Huang et al., Methods of Molecular Biology, 2008, Vol. 444, pp. 203-208), etc., which are incorporated herein by reference. By way of non-limiting example, peripheral blood can be collected in EDTA tubes, which can then be fractionated into plasma, leukocytes, and erythrocytes by centrifugation. DNA can be extracted from a cell-free plasma fraction (e.g., 0.5 to 2.0 mL) using a QIAamp DNA Blood Mini Kit (Qiagen, Valencia, CA) or similar kit according to the manufacturer's protocol.

[0158] In one aspect, the sample of lymphocytes used to generate the clonotype profile is large enough so that substantially every T cell or B cell with a different clonotype is present therein. In one embodiment, the sample taken contains every clonotype in a population with a frequency of 0.001% or more with a probability of 99%. In another embodiment, the sample taken contains every clonotype in a population with a frequency of 0.0001% or more with a probability of 99%. In another embodiment, the sample taken contains every clonotype in a population with a frequency of 0.00001% or more with a probability of 99%. In other embodiments, the sample taken contains every clonotype in a population with a frequency of 0.001% or more with a probability of 95%. In another embodiment, the sample taken contains every clonotype in a population with a frequency of 0.0001% or more with a probability of 95%. In another embodiment, the sample taken contains every clonotype in a population with a frequency of 0.00001% or more with a probability of 95%. In another embodiment, the sample taken contains every clonotype in a population with a frequency of 0.00001% or more with a probability of 95%. In yet another embodiment, the sample of B cells or T cells contains at least half a million cells, and in another embodiment, such a sample contains at least one million cells.

[0159] Whenever the source of material from which a sample is obtained is insufficient, such as clinical research samples, etc., DNA from the material can be amplified by an unbiased technique, such as whole genome amplification (WGA), multiple displacement amplification (MDA), or a similar technique, for example Hawkins et al, Curr. Opin. Biotech., 13:65-67 (2002); Dean et al, Genome Research, 11:1095-1099 (2001); Wang et al, Nucleic Acids Research, 32:e76 (2004); Hosono et al, Genome Research, 13:954-964 (2003) (Hosono et al., Genome Research, 2003, Vol. 13, pp. 954-964) etc.

[0160] Blood samples are of particular interest and can be obtained using conventional techniques such as Innis et al, editors, PCR Protocols (Academic Press, 1990). For example, leukocytes can be isolated from a blood sample using conventional techniques such as the Rosette Sep kit (Stem Cell Technologies, Vancouver, Canada). The volume of the blood sample can range from 100 μL to 10 mL; in one aspect, the blood sample volume ranges from 100 μL to 2 mL. DNA and / or RNA can then be extracted from such blood samples using conventional techniques suitable for use in the methods of the present invention, such as the DNeasy Blood & Tissue Kit (Qiagen, Valencia, CA). Optionally, subsets of leukocytes (e.g., lymphocytes) may be further separated using conventional techniques such as fluorescence activated cell sorting (FACS) (Becton Dickinson, San Jose, CA), magnetic activated cell sorting (MACS) (Miltenyi Biotec, Auburn, CA), etc.

[0161] Since there is recognition recombination in the DNA of each individual's adaptive immune cells and their related RNA transcripts, RNA or DNA can be sequenced by the method of the present invention provided. The recombinant sequence from a T cell or B cell is called a clonotype, and the recombinant sequence encodes a chain of a T cell receptor or an immunoglobulin molecule, or a portion thereof. The DNA or RNA may correspond to a sequence from a T cell receptor (TCR) gene or an immunoglobulin (Ig) gene encoding an antibody. For example, the DNA and RNA may correspond to a sequence encoding an α, β, γ or δ chain in a TCR. In most T cells, TCR is a heterodimer comprising an α chain and a β chain. The TCR α chain is generated by VJ recombination, while the β chain receptor is generated by V(D)J recombination. The TCR β chain in humans has 48 V segments, 2 D segments and 13 J segments. Several bases may be missing, and others (called N nucleotides and P nucleotides) are added to each of the two junctions. In a few T cells, TCR consists of γ and δ chains. TCRγ chains are generated by VJ recombination, while TCRδ chains are generated by V(D)J recombination (Kenneth Murphy, Paul Travers, and Mark Walport, Janeway's Immunology 7th edition, Garland Science, 2007, which is incorporated herein by reference in its entirety).

[0162] The DNA and RNA analyzed in the method of the present invention may correspond to some sequences, which encode heavy chain immunoglobulins (IgH) with constant regions (α, δ, ε, γ or μ) or light chain immunoglobulins (IgK or IgL) with constant regions λ or K. Each antibody has two identical light chains and two identical heavy chains. Each chain consists of a constant (C) region and a variable region. For the heavy chain, the variable region consists of a variable (V) segment, a diversity (D) segment and a connecting (J) segment. There are several different sequences in the genome to encode each type of these segments. Specific VDJ recombination events occur during the development of B cells, marking the cell to produce a specific heavy chain. The diversity of light chains is generated in a similar manner, but there is no D region, so there is only VJ recombination. Somatic mutations often occur near the recombination site, resulting in the addition or deletion of several nucleotides, further increasing the diversity of heavy and light chains produced by B cells. Then, different heavy and light chains cause the possible diversity of antibodies produced by B cells. The variable regions of the heavy and light chains contribute to the formation of antigen recognition (or binding) regions or sites. This diversity is also generated by somatic hypermutation, which can occur after a specific response to some epitopes is generated.

[0163] As described above, according to the present invention, primers can be selected to generate amplicons, which include portions of cell-free nucleic acids from lymphocytes or from tissues (such as blood) of recombinant nucleic acids. These portions may be referred to herein as "somatic rearrangement regions". Somatic rearrangement regions may include nucleic acids from developing or fully developed lymphocytes, in which the rearrangement of immune genes is not yet complete, so that molecules with, for example, complete V(D)J regions are not formed. Exemplary incomplete somatic rearrangement regions include incomplete IgH molecules (e.g., molecules containing only DJ regions), incomplete TCR molecules (e.g., molecules containing only DJ regions), and inactive IgK (e.g., containing Kde-V regions).

[0164] Amplification of nucleic acid populations

[0165] In some embodiments, the primer sequences of the first set of primers and the second set of primers can be selected according to conventional multiplex polymerase chain reaction (PCR). For example, guidance for selecting primers and performing multiplex PCR on nucleic acids encoding various immune receptor chains can be found in the following references, which are incorporated by reference: Faham and Willis, US patents 8,236,503 and 8,628,927; Morley, US patent 5,296,351; Gorski, US patent 5,837,447; Dau, US patent 6,087,096; Van Dongen et al, US patent publication 2006 / 0234234; European patent publication EP 1544308B1 (European Patent Publication EP1544308B1); Van Dongen et al, Leukemia, 17: 2257-2317 (2003) (Van Dongen et al, Leukemia, 2003, Vol. 17, pp. 2257-2317), etc. Guidance for multiplex PCR can be found in Henegariu et al, BioTechniques, 23: 504-511 (1997) (Henegariu et al, BioTechniques, 1997, Vol. 23, pp. 504-511), and similar references. In some embodiments, primers can be selected so that the frequency of amplified sequences in the final product is substantially the same as the frequency of sequences in the initial reaction mixture. Such primer selection can include selection of primer length, primer binding site, and primer concentration. As described above, the degree of multiplexing can vary greatly depending on the method selected for generating sequence reads and attaching sequence tags.

[0166] In some embodiments, the step of amplifying the target nucleic acid includes linear amplification of the target nucleic acid, which can be performed, for example, by repeating the cycle of annealing a set of primers (e.g., a first set of "upstream" or "forward" primers), extending the primers, and allowing the extended chains to melt from the template, so that the amount of the extended chain is amplified as a linear function of the number of cycles. In other words, the amplification step includes replicating the target polynucleotide (i.e., at least one chain of the target polynucleotide) by repeatedly extending a set of primers. In some embodiments, a step of removing the unextended primers can be performed after completing such a single or repeated extension in one direction, and a single or repeated extension can be performed on another set of primers (e.g., a second set of "downstream" or "reverse" primers) in the other direction.

[0167] Depending on the number and type of immune receptor chain nucleic acids amplified in the analysis, the number of primers in the first primer set and the second primer set may vary greatly. In some embodiments, common primers for various chains can be used. In other embodiments, specific primers can be designed for each different target polynucleotide to be amplified. Typically, the first primer set and the second primer set each contain multiple primers. In some embodiments, the multiple primers in the first primer set or the second primer set are at least 50 primers; in other embodiments, the multiple primers in the first primer set or the second primer set are at least 100 primers; in other embodiments, the multiple primers in the first primer set or the second primer set are at least 150 primers; in other embodiments, the multiple primers in the first primer set or the second primer set are at least 200 primers; in other embodiments, the multiple primers in the first primer set or the second primer set are at least 250 primers. The number of primers in the first group may be the same as or different from the number of primers in the second group.

[0168] In some embodiments, the first and second sets of primers are selected such that the length of the clonotype is at least 30 nucleotides; in other embodiments, the first and second sets of primers are selected such that the length of the clonotype is in the range of 30 to 500 nucleotides; in other embodiments, the first and second sets of primers are selected such that the length of the clonotype is in the range of 30 to 400 nucleotides; in other embodiments, the first and second sets of primers are selected such that the length of the clonotype is in the range of 30 to 300 nucleotides; in other embodiments, the first and second sets of primers are selected such that the length of the clonotype is in the range of 30 to 200 nucleotides.

[0169] Exemplary PCR amplification protocols can be found in van Dongen et al, Leukemia, 17:2257-2317 (2003) or van Dongen et al, US patent publication 2006 / 0234234, which are incorporated by reference. Briefly, an exemplary protocol is as follows: Reaction buffer: ABI Buffer II or ABI Gold Buffer (Life Technologies, San Diego, CA); 50 μL final reaction volume; 100 ng sample DNA; 10 pmol of each primer (adjusted as described below to balance amplification); 200 μM dNTPs at a final concentration; 1.5 mM MgCl2 at a final concentration (optimized according to the target sequence and polymerase); Taq polymerase (1-2 U / tube); Cycling conditions: pre-activation at 95°C for 7 min; annealing at 60°C; Cycling time: 30 s denaturation; 30 s annealing; 30 s extension. Polymerases that can be used for amplification in the methods of the present invention are commercially available, including, for example, Taq polymerase, AccuPrime polymerase, or Pfu. The polymerase to be used can be selected based on whether fidelity or efficiency is preferred.

[0170] Real-time PCR, picogreen staining, nanofluidics electrophoresis (eg LabChip) or UV absorbance measurements can be used in an initial step to estimate the functional amount of amplifiable material in the sample.

[0171] In one aspect, multiple amplification of the present invention is performed so that the relative amount of the sequence in the starting population is substantially the same as the relative amount of the sequence in the amplified population (or amplicon). That is, multiple amplification is performed when the amplification bias is lowest in the sample population member sequence. In one embodiment, if each relative amount in the amplicon is within five times of its value in the starting sample, such relative amounts are substantially the same. In another embodiment, if each relative amount in the amplicon is within twice of its value in the starting sample, such relative amounts are substantially the same. As discussed more fully below, conventional techniques can be used to detect and correct the amplification bias in PCR, so that a group of PCR primers can be selected for a predetermined group library, and the PCR primer set can provide a non-biased amplification of any sample.

[0172] In some embodiments, amplification bias can be avoided by performing a two-stage amplification (e.g., as described in Faham and Willis cited above), wherein a small number of amplification cycles (e.g., 2-5 or 2-10 or 2-15 cycles) are completed in the first stage (or primary stage), and the primers used in this stage have tails that are not complementary to the target sequence. The tail contains primer binding sites added to the ends of the primary amplicon sequence, so that these sites can be used in the second stage amplification using only a single forward primer and a single reverse primer, thereby eliminating the main cause of amplification bias. Before the start of the second stage amplification, the unextended primers of the first stage are removed from the reaction mixture or inactivated. In some embodiments, the number of cycles of the primary PCR will be as few as (e.g., 2-10) sufficient to minimize differential amplification caused by different primers. Secondary amplification is then performed with a pair of primers, which eliminates the source of differential amplification. In some embodiments, a small portion (e.g., 1%) of the reaction volume of the primary PCR is added directly to the secondary PCR reaction mixture. In some embodiments, the total number of cycles allocated to the first stage amplification and the second stage amplification is at least thirty-five.

[0173] In some embodiments, internal standards can be mixed with sample recombinant nucleic acids and amplified in the same reaction. Internal standards are nucleic acids with known sequences and concentrations. For example, they may be cloned copies of natural nucleic acids that encode portions of immune receptor chains; or, they may be synthetic nucleic acids. In some embodiments, the length and base composition of the internal standards are selected to represent specific immune receptor chains in amplification. By monitoring the relative concentration changes of the internal standards after amplification, amplification bias can be detected, and the conditions for non-biased amplification can be determined. For example, the primer length, position, and concentration can be changed to minimize the bias of the amplified product. In some embodiments, multiple internal standards are used in the reaction; in some embodiments, 2 to 50 different internal standards are used in the reaction; in other embodiments, 2 to 25 different internal standards are used in the reaction; and in some embodiments, 2 to 10 different internal standards are used in the reaction. In some embodiments, amplification bias is determined by measuring the relative frequency of sequences of different target nucleotides (e.g., all clonal types or selected clonal types or internal standards) in the amplified product. In other embodiments, the presence or level of amplification bias can be determined by performing real-time quantitative PCR on selected nucleic acids (such as two or more internal standards). Internal standards can also be used to quantify the number of different clonotypes in the original sample.Techniques for such molecular counting are well known, such as Brenner et al, US patent 7,537,897, which is incorporated herein by reference.

[0174] Generating sequence reads

[0175] The methods of the present invention may use any high throughput technology for sequencing nucleic acids. Preferably, such technology is capable of generating a large amount of sequence data in an economically efficient manner from which at least 1000 clonotypes can be determined, preferably at least 10,000 to 1,000,000 clonotypes can be determined. DNA sequencing technologies include: classical dideoxy sequencing reactions (Sanger method, which uses labeled terminators or primers and gel separation in plates or capillaries), sequencing by synthesis using labeled nucleotides that are reversibly terminated, pyrosequencing, 454 sequencing, allele-specific hybridization to labeled oligonucleotide probe libraries, sequencing by synthesis using allele-specific hybridization to labeled clone libraries followed by ligation, real-time monitoring of the incorporation of labeled nucleotides during the polymerization step, polymerase clone sequencing, and SOLiD sequencing. Sequencing of isolated molecules has recently been demonstrated by continuous or single extension reactions using polymerases and ligases, and single or continuous differential hybridization using probe libraries. These reactions have been performed on many parallel clone sequences, including demonstrations in commercial applications that currently exceed 100 million parallel sequences. These sequencing methods can therefore be used to study T cell receptor (TCR) and / or B cell receptor (BCR) repertoires. In one aspect of the invention, a high throughput sequencing method is employed, the method comprising the step of spatially isolating each molecule on a solid surface where each molecule is sequenced in parallel.Such solid surfaces may include non-porous surfaces (such as those used in Solexa sequencing, e.g., Bentley et al, Nature, 456:53-59 (2008) or whole genome sequencing, e.g., Drmanac et al, Science, 327:78-81 (2010)), well arrays that may include beads or particles bound to templates (such as those used in 454 sequencing, e.g., Margulies et al. al, Nature, 437:376-380 (2005)(Margulies et al., Nature, 2005, Vol. 437, pp. 376-380), or templates used in ion torrent sequencing (U.S. Patent Publication Nos. 2010 / 0137143 or 2010 / 0304982)), micromechanical membranes (such as those used in SMRT sequencing, for example Eid et al, Science, 323:133-138 (2009)(Eid et al., Science, 2009, Vol. 323, pp. 133-138)), or microbead arrays (such as those used in SOLiD sequencing or polymerase cloning sequencing, for example Kim et al, Science, 316:1481-1414 (2007)(Kim et al., Science, 2007, Vol. 316, pp. 1481-1414)). In another aspect, such methods include amplifying the separated molecules before or after spatially separating the molecules on a solid surface. Pre-amplification may include emulsion-based amplification (such as emulsion PCR) or rolling circle amplification.

[0176] Of particular interest are sequencing-by-synthesis methods using reversible terminators, such as Solexa-based sequencing, in which individual template molecules are spatially separated on a solid surface and then amplified in parallel by bridge PCR to form separate clonal populations or clusters, which are then sequenced, such as described in Bentley et al. (cited above) and the manufacturer's instructions (e.g., TruSeq TM Sample preparation kit and data sheet (TruSeq TMThe invention is described in the Sample Preparation Kit and Data Sheet, obtained from Illumina, Inc., San Diego, CA in 2010; and in the following references: U.S. Patents 6,090,592, 6,300,070, 7,115,400, and EP0972081B1; these references are incorporated by reference. In one embodiment, the cluster density formed by each molecule arranged and amplified on the solid surface is at least 10 5 Cluster / cm 2 ; or a density of at least 5×10 5 Cluster / cm 2 ; or a density of at least 10 6 Cluster / cm 2 . Solexa-based sequencing is also capable of generating two sequence reads from the same target sequence (or template) in a cluster, the two sequence reads being from opposite ends of the target sequence. In some embodiments, such sequence read pairs can be combined and treated as a single sequence read in subsequent analysis, or such sequence read pairs can be treated separately, but it should be taken into account that they are from the same cluster. Sometimes, sequence read pairs from the same template are referred to as "pairing", and the process of sequencing from both ends of a template is referred to as "bidirectional" sequencing. In some embodiments, the step of sequencing by synthesis using reversibly terminated labeled nucleotides includes: generating a single sequence read for each cluster or clonal population of the template, and generating multiple sequence reads (including but not limited to pairing) for each cluster or clonal population of the template. In another embodiment, when multiple sequence reads are generated for each cluster or clonal population of the template, such multiple sequence reads can be combined to form a single valid sequence read for subsequent analysis (such as a merging step).

[0177] In one aspect, a sequence-based clonotype profile of an individual sample is obtained using the following steps: (a) obtaining a nucleic acid sample from a T cell and / or B cell of the individual; (b) spatially separating molecules derived from the nucleic acid sample, each molecule comprising at least one template generated from nucleic acid in the sample, the template comprising a somatic rearrangement region or a portion of the region, each of the molecules being capable of generating at least one sequence read; (c) sequencing the spatially separated molecules; and (d) determining the abundance of different sequences of nucleic acid molecules in the nucleic acid sample to generate a clonotype profile. In one embodiment, each somatic rearrangement region comprises a V region and a J region. In another embodiment, the sequencing step comprises generating a plurality of sequence reads for each determined clonotype. In other embodiments, the sequencing step comprises combining information or data from a plurality of sequence reads to form each clonotype. In some embodiments, this combining step can be performed by merging sequence reads, as described by Faham and Willis in U.S. Pat. No. 8,628,927 (which is hereby incorporated by reference for use in the present teachings), or by using sequence tags, as described by Faham et al. in U.S. Patent Publication No. 2013 / 0236895A1 (which is hereby incorporated by reference for use in the present teachings). In another embodiment, the sequencing step includes bidirectional sequencing of each of the spatially separated molecules, generating at least one forward sequence read and at least one reverse sequence read.

[0178] Further to the latter embodiment, the at least one forward sequence read and the at least one reverse sequence read have an overlapping region, so that the bases of the overlapping region can be determined by the reverse complementary relationship between such sequence reads. In yet another embodiment, each somatic rearrangement region includes a V region and a J region, and the sequencing step also includes determining the sequence of each of the nucleic acid molecules by one or more forward sequence reads and at least one reverse sequence read of each nucleic acid molecule, wherein the reverse sequence read starts at a position in the J region and extends in the direction of its associated V region. In another embodiment, each molecule comprises a nucleic acid selected from the following molecules: a complete IgH molecule, an incomplete IgH molecule, a complete IgK molecule, a complete IgK inactive molecule, a TCRβ molecule, a TCRγ molecule, a complete TCRδ molecule, and an incomplete TCRδ molecule. In another embodiment, the sequencing step includes generating sequence reads having a monotonically decreasing quality score. In another embodiment, the above method comprises the following steps: (a) obtaining a nucleic acid sample from an individual's T cells and / or B cells; (b) spatially separating molecules derived from such nucleic acid sample, wherein each molecule comprises a nested group of templates, each template being generated from nucleic acid in the sample and comprising a somatic rearrangement region or a portion of such region, wherein each nested group is capable of generating a plurality of sequence reads, each sequence read extending in the same direction and starting from a different position on the nucleic acid where the nested group is generated; (c) sequencing each of the spatially separated molecules; and (d) determining the abundance of different sequences of nucleic acid molecules in the nucleic acid sample to generate a clonal profile. In one embodiment, the sequencing step comprises generating a plurality of sequence reads for each nested group. In another embodiment, each somatic rearrangement region comprises a V region and a J region, and each of the plurality of sequence reads starts at a different position in the V region and extends in the direction of its associated J region.

[0179] In one aspect, for each sample from an individual, the sequencing technology used in the method of the present invention generates sequences of at least 1000 clonotypes per run; in another aspect, such technology generates sequences of at least 10,000 clonotypes per run; in another aspect, such technology generates sequences of at least 100,000 clonotypes per run; in another aspect, such technology generates sequences of at least 500,000 clonotypes per run; in another aspect, such technology generates sequences of at least 1,000,000 clonotypes per run. In another aspect, for each individual sample, such technology generates sequences of between 100,000 and 1,000,000 clonotypes per run. In each of the above aspects, each clonotype per run is determined by at least 10 sequence reads.

[0180] The sequencing technology used in the provided methods of the present invention can produce about 30bp, about 40bp, about 50bp, about 60bp, about 70bp, about 80bp, about 90bp, about 100bp, about 110bp, about 120bp, about 150bp, about 200bp, about 250bp, about 300bp, about 350bp, about 400bp, about 450bp, about 500bp, about 550bp, or about 600bp per read.

[0181] Determining clonotypes from sequence data

[0182] In some embodiments of the present invention, sequence tags are used to determine clonotypes, and in other embodiments, sequence tags are used in combination with a sequence read merging step to determine clonotypes. In embodiments where a single unique sequence tag is attached to substantially each different target polynucleotide, the sequence tag can be used directly to determine the clonotype. In such embodiments, the sequence reads are first grouped based on the sequence tags of the sequence reads to determine the clonotypes of the sample. This grouping can be accomplished by conventional sequence alignment methods. Guidance for selecting alignment methods can be found in Batzoglou, Briefings in Bioinformatics, 6: 6-22 (2005), which is incorporated by reference. After the sequence reads are grouped corresponding to the unique sequence tags, the sequences of the associated clonotypes can be analyzed to determine the sequences of the clonotypes from the sample. Figure 4A An exemplary alignment and method for determining a sequence (SEQ ID NO:2) of a clonotype associated with a unique sequence tag is shown. In this example, eleven sequence reads are aligned by their respective sequence tags 4302, and after alignment, the nucleotides at each position of the clonotype portion 4304 of the sequence reads (shown as 1, 2, 3, 4, ... n) are compared. For example, the nucleotides at position 6 (4306) are t, t, g, t, t, t, t, t, c, t; that is, nine base calls are t, one is "g" (4308) (SEQ ID NO:3), and one is "c" (4310) (SEQ ID NO:4). In one embodiment, the correct base call for the clonotype sequence at a position is the type of base that is in the majority. In the example of position 6 (4306), the base call is "t" because t is the nucleotide in the majority of sequence reads at that position. In other embodiments, other factors may be considered to determine the correct base call of the clonotype sequence, such as the quality score of the base call of the sequence read, the types of adjacent bases, etc. After the clonotypes are determined as described above, a clonotype profile may be assembled, which includes the abundance or frequency of each different clonotype in the sample.

[0183] In some embodiments, more than one extension step may be performed using primers containing sequence tags, thereby increasing the proportion of target polynucleotides in the sample that are labeled with sequence tags before amplification. In such embodiments, performing more than one extension step in the presence of primers containing sequence tags causes the target polynucleotide and / or its copies to be labeled with multiple different sequence tags. The number of different sequence tags depends on the number of extension steps performed in the presence of primers containing sequence tags, the efficiency of the amplification reaction, whether both the forward primer and the reverse primer have sequence tags, and the like. In some such embodiments, the number of different sequence tags is in the range of 2 to 15, or in the range of 2 to 10, or in the range of 2 to 5. In some such embodiments, after amplification, the copies of each target polynucleotide in the sample can be divided into multiple groups or subsets, wherein the members of each group or subset are labeled with the same sequence tag, and the members of each different group or subset are labeled with different sequence tags; that is, the members of the same group have the same sequence tag, and the members of different groups have different sequence tags. In other words, after amplification, in some embodiments, each copy of the target polynucleotide in the sample will be labeled with one of two different sequence tags; or in other embodiments, each copy of the target polynucleotide in the sample will be labeled with one of three different sequence tags; or in other embodiments, the copies of the target polynucleotide in the sample will be labeled with one of four different sequence tags; etc. In these embodiments, clonotypes can be determined by performing a merging step after sequence tag alignment, wherein the merging step regards sequence reads within a group as originating from the same parental sequence based on the probability that the common origin is true, which varies with error rate, relative frequency, etc. Figure 4B Sequence reads from such an embodiment are shown. In one approach, sequence reads are first grouped by common sequence tags 4402, shown in the figure as three groups 4420, 4422, and 4424. In some embodiments, the sequences within each group 4404 are analyzed to determine a consensus sequence for the group; for example, as described above, the base at each nucleotide position may be called the majority base or the highest frequency base, etc. The consensus sequences of the groups may then be merged with each other to determine a clonotype.

[0184] In some embodiments, the above aspects of the present invention can be implemented in a method for analyzing almost all nucleic acid populations in a sample. This method may include the following steps: (a) obtaining a sample comprising a nucleic acid population; (b) attaching a sequence tag to the nucleic acid of the population to form a tag-nucleic acid conjugate, wherein at least one nucleic acid of the population or a copy thereof is attached with a different sequence tag; (c) amplifying the tag-nucleic acid conjugate; (d) sequencing the tag-nucleic acid conjugate to produce a sequence read, the sequence read having an error rate and comprising a nucleic acid sequence and a tag sequence; (e) aligning sequence reads with similar tag sequences to form a sequence read group with the same sequence tag; (f) merging the sequence reads of each group to determine the sequence of the nucleic acid, wherein each time the sequence read group has at least 95% probability of being different, the sequence read group is merged into a different sequence; and (g) determining the sequence profile of the population by determining the level of the sequence. When this method is applied to analyze a recombinant nucleic acid population, the method can be implemented by the following steps: (a) obtaining a sample containing T cells and / or B cells and / or cell-free DNA from an individual; (b) attaching sequence tags to recombinant nucleic acid molecules of T cell receptor genes or immunoglobulin genes from the sample to form tag-nucleic acid conjugates, wherein at least one recombinant nucleic acid or a copy thereof from the sample is attached with a different sequence tag; (c) amplifying the tag-nucleic acid conjugate; (d) sequencing the sample of the tag-nucleic acid conjugate to provide sequence reads, wherein the sequence reads have an error rate and include a tag sequence and a recombinant nucleic acid sequence; (e) aligning sequence reads with similar tag sequences to form a group of sequence reads with the same sequence tag; (f) merging sequence reads of each group to determine a clonotype, wherein each time the sequence read group has at least a 95% probability of being different, the sequence read group is merged into a different sequence; and (g) determining the clonotype profile of the sample by determining the level of the clonotype.

[0185] In the above embodiments and other embodiments disclosed herein, the step of sequencing the tag nucleic acid conjugate includes sequencing a sample of the tag nucleic acid conjugate from the amplicon. Typically, such a sample is a representative sample, wherein the relative frequency of the target polynucleotide in the original sample (i.e., tissue sample, blood sample, etc.) is maintained in the sample of the tag-nucleic acid conjugate from the amplification reaction product. In some embodiments of analyzing a population of recombinant nucleic acids encoding immune receptor molecules, the sample of the tag-nucleic acid conjugate contains at least 10 4 tag-nucleic acid conjugates; in other embodiments, such samples contain at least 10 5 tag-nucleic acid conjugates; in other embodiments, such samples contain at least 10 6 tag-nucleic acid conjugates; in other embodiments, such samples contain at least 10 7 A tag-nucleic acid conjugate.

[0186] Merging sequence reads

[0187] In embodiments where multiple sequence tags are attached to the original recombinant nucleic acid or a copy thereof, a step of merging sequence reads (or common sequence reads for each group) may be performed to determine clonotypes. If the sequencing technology is error-free, it would be trivial to reduce a set of sequence reads for a given sample to a set of different clonotypes and record the number of sequence reads for each clonotype. However, in the presence of sequencing errors, each true clonotype is surrounded by a "cloud" of sequence reads with different numbers of errors associated with its sequence. As the distance of the sequencing error "cloud" from the clonotype in sequence space increases, the density of the "cloud" decreases. There are a variety of algorithms available for converting sequence reads into clonotypes. In one aspect, merging sequence reads (i.e., merging candidate clonotypes that have been determined to have one or more sequencing errors) depends on at least three factors: the number of sequences obtained for each clonotype being compared, the number of bases that are not identical in the sequence, and the sequencing quality score at the position of the sequence inconsistency. In some embodiments, a likelihood ratio may be constructed and evaluated based on an expected error rate and a binomial distribution of errors. For example, there are two clonotypes, one of which has 150 sequence reads and the other has 2 sequence reads. There is one difference between the two clonotypes in the area of ​​poor sequencing quality. Because the two are likely to be generated by sequencing errors, the two can be merged. On the other hand, there are two clonotypes, one of which has 100 sequence reads and the other has 50 sequence reads. There are two differences between the two. Because it is considered that the two clonotypes are unlikely to be generated by sequencing errors, the two are not merged. In some embodiments, the following algorithm can be used to determine clonotypes from sequence reads. Figure 5A Some of these concepts are exemplified in FIG. 1 . In some embodiments of the merging step, the sequence reads are first converted into candidate clonotypes. This conversion depends on the sequencing platform used. For platforms that produce long sequence reads with high quality scores, the sequence reads or a portion thereof can be directly used as candidate clonotypes. For platforms that produce shorter sequence reads with low quality scores, some alignment and grouping steps are required to convert a group of related sequence reads into candidate clonotypes. For example, in some embodiments, for a Solexa-based platform, candidate clonotypes are generated from a collection of paired sequence reads from multiple clusters (e.g., 10 or more clusters) as described above.

[0188] like Figure 5AAs shown, the frequencies of candidate clonotypes may be plotted in sequence space, which is simplified to one dimension (horizontal axis) for ease of illustration. The vertical axis represents the magnitude of the frequency of each candidate clonotype, or log(read count), or some similar measure. In the figure, candidate clonotypes are represented by various symbols 530. According to one embodiment of the present invention, whether to merge two candidate clonotypes depends on their respective frequencies or read counts (as described above), the number of base differences between the two (the more differences, the less likely they are to be merged), and the quality scores of the bases at the positions where their respective sequences differ (the higher the quality score, the less likely they are to be merged). Candidate clonotypes may be considered in the order of their respective frequencies. Figure 5A Candidate clonotype 1 (532), candidate clonotype 7 (534), and candidate clonotype 11 (536) are shown as the three candidate clonotypes with the highest three frequencies. Associated with each such candidate clonotype are other candidate clonotypes that are close in sequence but less frequent, such as (i) for candidate clonotype 1 (532), candidate clonotype 2 (538) and candidate clonotypes 3, 4, 5, and 6 surrounded by cone (540) are associated with it; (ii) for candidate clonotype 7 (534), candidate clonotypes 8, 9, and 10 surrounded by cone (542) are associated with it; (iii) for candidate clonotype 11, candidate clonotype 12 surrounded by cone (544) is associated with it. The cone represents the boundary of possibility within which the lower frequency candidate clonotypes will merge with one of the higher frequency candidate clonotypes 1, 7, or 11. This possibility boundary is a function of the frequencies of nearby candidate clonotypes (near 1 are 3, 4, 5, and 6; near 7 are 8, 9, and 10; near 11 is 12) and their distance in sequence space from the corresponding higher frequency candidate clonotypes. Candidate clonotype 2 (538) is outside the cone (540); therefore, it is not merged with candidate clonotype 1 (532). Again, the (merged) possibility boundary is shown as a cone because candidate clonotypes with higher frequencies are more likely to be truly different clonotypes than candidate clonotypes with lower frequencies, while multiple differences at lower frequencies are more likely to be errors than multiple differences at higher frequencies.

[0189] The cloud of sequence reads around each candidate clonotype can be modeled using a binomial distribution and a simple model of the probability of a single base error. The latter error model can be inferred by self-consistency and convergence from mapping V and J segments or from the clonotype finding algorithm itself. The model is constructed to obtain the probability that a given "cloud" sequence Y with read count C2 and E errors (relative to sequence X) is part of a true clonotype sequence X with ideal read count C1 under the null model, that is, X is the only true clonotype in this region of sequence space. The decision whether to merge sequence Y into clonotype X is made based on the parameters C1, C2, and E. For any given C1 and E, the maximum value C2 is pre-calculated to determine whether to merge sequence Y. The maximum value of C2 is chosen so that under the null hypothesis that Y is part of clonotype X, after integrating all possible sequences Y with errors E near sequence X, the probability of not being able to merge Y is less than a certain P value. The P value can control the behavior of the algorithm and make merging more or less permissive.

[0190] If sequence Y cannot be merged into clonotype X because the read count of sequence Y is higher than the threshold C2 for merging into clonotype X, then sequence Y becomes the one used to form another clonotype (such as Figure 5A An algorithm implementing this principle will also ensure that any other sequences Y2, Y3, etc. that are "closer" to sequence Y (which is considered independent of X) will not be clustered around X. The concept of "close" includes both the error counts with respect to Y and X as well as the absolute read counts of X and Y, i.e., it is modeled in the same way as the model described above for the cloud of erroneous sequences around clonotype X. In this way, if "cloud" sequences happen to be "close" to more than one clonotype, they can be correctly attributed to their correct clonotypes. Therefore, see Figure 5A , if candidate clonotype 2 is considered truly different from candidate clonotype 1 (532), then a dedicated program or sub-algorithm will provide rules to determine which of candidate clonotypes 1 (532) and 2 (538), candidate clonotypes 4 and 5, and 1 and 2 (if any) should be merged.

[0191] In one embodiment, the algorithm starts with the sequence X with the highest read count and proceeds in a top-down manner. This sequence forms the first clonotype. If the count of a neighboring sequence is below a pre-calculated threshold (see above), it is merged into this clonotype, and if the read of a neighboring sequence is above the threshold or "closer" to another unmerged sequence, it is left alone. After searching all neighboring sequences within the maximum error count, the process of merging reads into clonotype X stops. Its reads and all reads that have been merged into it are interpreted and removed from the list of reads that can be used to generate other clonotypes. Then turn to the next sequence with the highest read count. As described above, neighboring reads are merged into this clonotype, and this process continues until there are no more sequences with read counts above a given threshold, for example until all sequences with counts greater than 1 have been used to form clonotypes.

[0192] As described above, in another embodiment of the above algorithm, a further test may be added to determine whether to merge candidate sequence Y into existing clonotype X, which test takes into account the quality scores of the related sequence reads. In the case where sequence Y and X are different, the average quality score of sequence Y is determined (the average of all reads with sequence Y is taken). If the average score is above a predetermined value, then the difference is more likely to indicate that the sequence is a truly different clonotype and should not be merged; if the average score is below the predetermined value, then sequence Y is more likely to be generated by a sequencing error and should therefore be merged into X.

[0193] The successful implementation of the algorithm for merging candidate clonotypes described above depends on whether one can efficiently find all sequences from some input sequence X that have a number of errors less than E (i.e., less than some sequence distance measure). This problem can be solved using a sequence tree. This implementation of a sequence tree has some unusual features because the nodes of the tree are not limited to single bases of the DNA sequence of the candidate clonotypes, e.g. Figure 5D These nodes can have arbitrarily long sequences to make more efficient use of computer memory.

[0194] All reads for a given sample are placed in a sequence tree. Each leaf node points to the read associated with it. Unique sequences for candidate clonotypes are obtained by traversing the sequence tree backwards from the leaf nodes to the root node. The first sequence is placed in a simple tree with one root node and one leaf node, which contains all the sequences of the reads. Then the sequences are added one by one. For each added sequence, a new branch is formed at the last point of common sequence between the read and the existing tree, or if the tree already contains this sequence, the read is added to the existing leaf node. After all the reads are placed in the tree, the tree can be easily used to achieve the following purposes: 1) Find the highest read count: sorting the leaf nodes by read count can find the leaf nodes (i.e., sequences) with the most reads, and then find the leaf nodes with a lower number of reads; 2) Find adjacent leaves: For any sequence, all paths in the tree with errors less than X relative to the sequence can be searched. The path starts at the root and branches into multiple separate paths to continue along the sequence tree. The current error count of each path is recorded as it extends along the sequence tree. When the error count exceeds the maximum allowed, the given path is terminated. In this way, most of the tree will be corrected as early as possible. This method can effectively find all paths (ie, all leaves) with errors in X from any given sequence.

[0195] Figure 5B The flowchart of illustrates the features of the above concepts in more detail. Sequence data is obtained by sequencing recombinant nucleic acids extracted from a sample of T cells or B cells, and a set of candidate clonotypes is obtained from the sequence data. In one aspect, each candidate clonotype comprises an NDN region and a portion of a V region and a J region. These sequences are placed in a data structure (550), which can be a sequence tree. In one embodiment, as part of generating a set of candidate clonotypes, a sequence tree can also be constructed for known V regions and known J regions (not in Figure 5B ). The sequence reads constituting the candidate clonotype can then be mapped to or aligned with these known sequences through the sequence tree to effectively determine the most likely known V and J sequences of the candidate clonotype. Figure 5B After generating candidate clonotypes, a data structure (such as a sequence tree) is constructed to be able to distinguish true clonotypes from candidate clonotypes containing experimental errors or measurement errors (such as sequencing errors). The candidate clonotype (HFCC) with the highest frequency of occurrence among the current candidate clonotypes is selected from the data structure (e.g., sequence tree). k )(552); In other words, HFCC k is the candidate clonotype with the highest copy number or the highest read count in cycle k. Next, the adjacent lower frequency candidate clonotypes (LFCCs) are identified (554); that is, the candidate clonotypes with a distance D kIn one aspect of the invention, this identification is performed using a sequence tree, which allows for efficient sequence comparison of relatively short (<300 bp) sequences.

[0196] In one embodiment, the comparison or sequence alignment is performed using a dynamic programming such as that disclosed by Gusfield (as mentioned above). In another embodiment, the dynamic programming is a banded dynamic programming, where sequences that differ from the selected HFCC by more than a predetermined distance are not considered, thereby speeding up the calculation. Candidate HFCCs may be compared based on many different criteria or characteristics. k and LFCC j In one aspect, as described above, candidate clonotypes are compared based on at least two characteristics: (i) frequency or read count and (ii) sequence difference. In another aspect, as described above, candidate clonotypes are compared based on at least three characteristics: (i) frequency or read count, (ii) sequence difference and (iii) a quality score or metric of the bases where the difference occurs. In one embodiment, the sequence difference comprises base substitutions; in another embodiment, the sequence difference comprises base substitutions, deletions and insertions. The latter embodiment is particularly suitable for generating sequence data by sequencing-by-synthesis methods that do not use terminators, such as 454 sequencers and Ion Torrent sequencers. Such sequencing methods distinguish between homopolymer stretches of different sizes by signal amplitude; therefore, the base calling process in such methods is prone to insertion and deletion errors because the degree of difference in signal level between homopolymers that differ by one nucleotide decreases sharply with increasing homopolymer size (i.e., 2-mers are easily distinguished from 3-mers, but 8-mers are difficult to distinguish from 9-mers). In one aspect, a function (referred to herein as a "combined likelihood function") may be used to compare the HFCC and the LFCC, such as P(HFCC) shown in decision block (558). k ,LFCC j,D,Q) function, depending on the quantities (i) to (iii) as described above. This function can take many different forms, but the P value generally changes with (i), (ii) and (iii) as follows: the P value preferably increases monotonically with the frequency of the HFCC and the ratio of the HFCC frequency to the LFCC frequency, so that the higher the ratio of the HFCC frequency to the LFCC frequency, the higher the likelihood LFCC to be merged into the HFCC. Similarly, the P value preferably decreases monotonically with the degree of sequence difference between the HFCC and the LFCC, so that the greater the difference between the HFCC and the LFCC (e.g., measured by the minimum number of substitutions, insertions or deletions that change one sequence to another), the lower the likelihood LFCC to be merged into the HFCC. Finally, the P value preferably decreases monotonically with the increase of the quality score of the position where the sequence difference between the HFCC and the LFCC occurs, so that the higher the quality score, the lower the likelihood LFCC to be merged into the HFCC.

[0197] When the sequence difference between HFCC and LFCC occurs at more than one position, the quality scores at different positions can be combined by a variety of different methods. In one embodiment, whenever there are multiple such differences, the multiple quality scores are represented by an average value, which can be an unweighted average or a weighted average. Figure 5C An exemplary function P is shown for calculating different quality values ​​for given sequence differences (curves a to e). Figure 5C As shown, whenever HFCC is at a level of about 200 read counts (570), if the quality score is determined by curve (a), then any LFCC (572) with a read count less than about 50 will be merged into HFCC. The independent variable D of function P is the sequence HFCC k and LFCC j D is a measure of the distance between , and its value may vary from cycle to cycle as the analysis proceeds. (The symbol "k" indicates that the value of a constant with a subscript "k" may depend on the computational cycle k.) In one embodiment, D = D k , so that its value is a function of the number of cycles. In another embodiment, D = D (HFCC frequency), so that its value is a function of the HFCC frequency and is independent of the number of cycles. For example, as the HFCC frequency decreases, the distance between the candidate items to be compared decreases. In one embodiment, D is HFCC k With LFCC j The Hamming distance between the two is ; however, other distance measures may also be used. In one embodiment, D k is a non-increasing function of k; and in another embodiment, D kis a decreasing function of k. In some embodiments, it is advantageous to reduce the magnitude of D as the cycle number increases or as the HFCC frequency decreases because the candidate clonotypes become less and less frequent as the calculation proceeds, most of which are singletons, making sequence distance (rather than frequency difference) the primary comparison. By decreasing D as the calculation proceeds, ineffective comparisons with distant low-frequency candidate clonotypes are reduced, thereby speeding up the calculation. Function P can be a complex expression, depending on the number of factors considered. Figure 5C Calculated values ​​for one embodiment of P are shown, which are related to the read count threshold for merging LFCC given HFCC read counts for different quality scores, as described above. Curves "a" to "e" represent the relationship of different quality scores (where curve "a" corresponds to the highest quality score).

[0198] Back to Figure 5B , if P <P k , then LFCC is not j With HFCC k Merge and select another LFCC (560). If P>P k , then LFCC j With HFCC k Merge (562), in which case another LFCC is selected (566), unless there are no more LFCCs left to be evaluated (564). If there are no more LFCCs to be evaluated (564), the current LFCC is removed from the data structure (e.g., sequence tree) k (including incorporation into HFCC k All LFCCs of (568). Such removal Figure 5D-5E 590. In the figure, the path (592) in the sequence tree (590) (indicated by the dotted line) corresponds to the HFCC (596), which is merged with the LFCC (598). After the merger, the segment of the path (592) in the shaded area (599) is removed from the sequence tree (590) to obtain Figure 5EThe reduced sequence tree (597) shown is used to find adjacent LFCCs (554) in subsequent calculations. After the removal, if the stopping criterion (570) is met, the determination of the clonotype is completed. In one embodiment, the stopping criterion (570) is whether the last non-singleton candidate clonotype (552) has been processed. In another embodiment, the stopping criterion (570) is whether the frequency or read count of the selected HFCC is lower than the frequency or read count of the HFCC corresponding to a single lymphocyte. In one aspect of the method of the present invention, the amplification step can cause each lymphocyte in the sample to be represented by multiple copies of the same clonotype; therefore, in one embodiment, the calculation is stopped whenever the number of read counts of the HFCC is lower than the number corresponding to a single lymphocyte. In some embodiments, the number of such read counts (or candidate clonotype copies) is at least 10; in another embodiment, this number is at least 20; in another embodiment, this number is at least 30; in another embodiment, this number is at least 40. If the stopping criterion is not met, the next HFCC is selected (572). Figure 5B The analysis steps summarized in the flowchart can be performed using any suitable programming language such as C, C++, Java, C#, Fortran, Pascal, etc.

[0199] According to one aspect of the present invention, the method for determining clonotypes and / or clonotype profiles comprises the following steps: (a) forming a data structure of recombinant immune molecules according to sequence reads obtained by high-throughput nucleic acid sequencing; (b) whenever the frequency of any candidate clonotype with a lower frequency is lower than a predetermined frequency value and the sequence difference between the candidate clonotype with the highest frequency and the candidate clonotype with the lower frequency is lower than a predetermined difference value, merging the candidate clonotype with the candidate clonotype with the highest frequency to form a clonotype; (c) removing the merged candidate clonotype from the data structure; and (d) repeating steps (b) and (c) until a clonotype profile is formed. In one embodiment, the data structure is a sequence tree.

[0200] According to another aspect of the present invention, the above method for determining clonotypes may be performed by the following steps: (a) providing a set of sequence reads by a repertoire of recombinant immune molecules each having a V region, an NDN region and a J region, wherein for each such molecule, at least one sequence read covers at least a portion of the NDN region of such molecule; (b) forming a sequence tree having leaves representing candidate clonotypes from the sequence reads covering at least a portion of the NDN region, each leaf and its corresponding candidate clonotype having a certain frequency; (c) whenever the frequency of any candidate clonotype with a lower frequency is lower than a predetermined frequency value and the sequence difference between the candidate clonotype with the highest frequency and the candidate clonotype with the lower frequency is lower than a predetermined difference value, merging the candidate clonotype with the highest frequency with such candidate clonotype with the lower frequency to form a clonotype having the sequence of the candidate clonotype with the highest frequency; (d) removing the leaf corresponding to the merged candidate clonotype from the sequence tree; and (e) repeating steps (c) and (d) until the highest frequency of the candidate clonotype with the lower frequency is lower than a predetermined stop value. In one embodiment, the forming step further comprises selecting the highest frequency candidate clonotype and identifying all of the lower frequency candidate clonotypes whose sequence difference with the selected highest frequency candidate clonotype is less than a predetermined difference value to form a merged subset. Thus, in such embodiments, the total number of LFCCs that must be compared by the merge operation can be limited (only LFCCs within the predetermined difference value are considered). Such values ​​are process inputs that depend on the specific application, such as the size of the repertoire, how much computational time is used, etc. As described above, the function for deciding whether to merge a HFCC with a LFCC can have a variety of forms. In a general aspect, for the merging step, such a function can have the following characteristics: it depends on the frequency of the HFCC, the frequency of the LFCC, the sequence difference between the HFCC and the LFCC (the sequence difference can be expressed as a conventional string difference measure, such as the Hamming distance), and the quality score of one or more nucleotide positions at which the HFCC and the LFCC differ; such that the function (i) increases monotonically with the increase of the ratio of the HFCC frequency to the LFCC frequency, (ii) decreases monotonically with the increase of the sequence difference between the HFCC and the LFCC, and (iii) decreases monotonically with the increase of the quality score of the one or more nucleotide positions. That is, with respect to characteristic (iii), the more certain it is that the HFCC and LFCC are different (eg, because there is a high confidence level in the base call), the less likely it is that the HFCC and LFCC will be merged.

[0201] In some embodiments, the merge likelihood function is selected so that whenever the sequence reads have at least a 95% chance of being different, such sequence reads are merged into different clonotypes (or target polynucleotides, such as recombinant nucleic acids); in other embodiments, the merge likelihood function is selected so that whenever the sequence reads have at least a 99% chance of being different, such sequence reads are merged into different clonotypes; in other embodiments, the merge likelihood function is selected so that whenever the sequence reads have at least a 99.9% chance of being different, such sequence reads are merged into different clonotypes. As described above, in some embodiments, the merge likelihood function depends on the error rate of the sequencing chemistry used, the number of inconsistent nucleotides in the compared sequence reads, and the relative frequency of the compared sequence reads; in another embodiment, the merge likelihood function depends on the error rate of the sequencing chemistry used, the number of inconsistent nucleotides in the compared sequence reads, the relative frequency of the compared sequence reads, and the quality score of the inconsistent nucleotides. In the above, the selection of the predetermined frequency value and the predetermined difference value is a design choice that depends on the specific application. Factors affecting such selection may include biological details, implementation speed, etc.

[0202] Monitoring Applications

[0203] In one aspect, the present invention relates to a method for monitoring minimal residual disease by determining the presence, absence and / or level of nucleic acids characterizing a disease or associated with a disease in a sample. In some embodiments, such nucleic acids are somatic recombinant nucleic acids, or clonotypes, which are associated with a precancerous condition or a cancer condition (such as a lymphoid or myeloproliferative disease), and which can be used to monitor the state of the disease or condition. Such nucleic acids, especially clonotypes, can be used to monitor minimal residual disease of cancer after treatment, wherein the results of such monitoring are a key factor in deciding whether to continue treatment, stop treatment, or otherwise improve treatment. For many malignant lymphoid and myeloid tumors, diagnostic tissue samples, such as peripheral blood samples or bone marrow samples, are obtained before treatment, and clonotype profiles ("diagnostic clonotype profiles") are generated using the samples. For lymphoid or myeloproliferative diseases, it is often unknown which immune receptor chains are associated with lymphoid or myeloid clones of the disease or condition before the diagnostic sample. Therefore, in current methods, multiple separate amplifications and sequencing of different recombinant nucleic acids encoding different candidate immune receptor chains must be performed in order to identify clonotypes associated with the patient's disease or condition. One or more clonotypes associated with the disease (i.e., "correlating clonotypes") are identified by the clonotype profiles generated from such amplification and sequencing efforts. Typically, the clonotype with the highest frequency in the clonotype profile is selected as the correlating clonotype. In one aspect of the invention, a larger scale of multiplex amplification is provided by performing a single reaction on several portions of a recombinant nucleic acid encoding a plurality of different immune receptor chains, greatly reducing the number of separate amplification and sequencing runs required to identify correlating clonotypes. In some embodiments, the "multiple" is in the range of 2 to 4 separate immune receptor chains; and in other embodiments, the "multiple" is in the range of 2 to 3 separate immune receptor chains. More specifically, in some embodiments, in the BCR chain, the following sequences are amplified by a single multiplex reaction: recombinant nucleic acids encoding IgH including at least a portion of the VDJ region, IgH including at least a portion of the DJ region, and IgK; and in other embodiments, in the TCR chain, the following sequences are amplified by a single multiplex reaction: TCRβ, TCRδ, and TCRγ.

[0204] After treatment, and preferably after complete remission of the cancer is achieved, the presence, absence or frequency of such correlating clonotypes or nucleic acids is periodically assessed to determine whether remission is maintained or tumor recurrence is present based on the presence or increased frequency of the correlating nucleic acid or clonotype (or involved clonotypes) in the post-treatment clonotype profile or nucleic acid profile. That is, after treatment, minimal residual lesions of the cancer are assessed based on the presence, absence or frequency of the correlating clonotypes or characteristic nucleic acids. As described above, when such correlating clonotypes are common in or correspond to rearranged receptor segments that lack sufficient diversity (such that non-cancerous cells can share the clonotype), the presence of such clonotypes in the post-treatment clonotype profile can produce a false positive indication of recurrence.

[0205] The method of the present invention is applicable to monitoring any proliferative disease, in which the rearranged nucleic acid encoding an immune receptor or a portion thereof can be used as a marker of the cells involved in the disease. In one aspect, the method of the present invention is applicable to lymphatic and myeloproliferative diseases. On the other hand, the method of the present invention is applicable to lymphoma and leukemia. On the other hand, the method of the present invention is applicable to monitoring follicular lymphoma, chronic lymphocytic leukemia (CLL), acute lymphocytic leukemia (ALL), chronic myeloid leukemia (CML), acute myeloid leukemia (AML), Hodgkin's lymphoma and non-Hodgkin's lymphoma, multiple myeloma (MM), monoclonal gammopathy of unknown significance (MGUS), mantle cell lymphoma (MCL), diffuse large B-cell lymphoma (DLBCL), myelodysplastic syndrome (MDS), T-cell lymphoma, etc. MRD (minimal residual lesions). In a specific embodiment, the method of the present invention is particularly applicable to monitoring the MRD (minimal residual lesions) of ALL, MM or DLBCL.

[0206] In some embodiments, a patient sample such as blood or bone marrow is subjected to diagnostic analysis to identify which of a plurality of immune receptor chains may contain a clonotype (i.e., a correlating clonotype) produced by a clone of the disease. Once the immune receptor chain of the correlating clonotype is determined, a subsequent monitoring analysis may be performed for that particular immune receptor chain. For example, in some embodiments, a diagnostic analysis may generate a sequence-based clonotype spectrum of multiple BCR chains (such as IgH (VDJ), IgH (DJ), and IgK) in the same reaction. If the correlating clonotype is an IgH (VDJ) chain, a subsequent monitoring analysis may only generate a clonotype spectrum of IgH (VDJ). In some embodiments, the sequencing depth in the diagnostic sample may be different from the sequencing depth in the monitoring sample. "Sequencing depth" refers to the total number of sequence reads analyzed for constructing a clonotype spectrum. For cancers such as leukemia and lymphoma, because the patient sample is first subjected to diagnostic analysis before treatment, the frequency or level of the correlating clonotype in the sample is usually high and easy to identify. For example, any clonotype whose frequency exceeds a predetermined level may be defined as a correlating clonotype. This predetermined level may be different when other patient indicators are used; and the predetermined level may generally be in the range of 2% to 5%; or in some embodiments, 5%. Thus, in some embodiments, the sequencing depth performed is that required to reliably detect clonal types present at a frequency of 1% or 2% or more. In some embodiments, the sequencing depth of the diagnostic sample produces at least 10,000 sequence reads; or in other embodiments, produces at least 100,000 sequence reads; in yet other embodiments, the sequencing depth of the diagnostic sample produces at least 10 6 In some embodiments, the sequencing depth of the monitoring sample is at least 100,000 sequence reads; in other embodiments, the sequencing depth of the monitoring sample is at least 10 6 sequence reads.

[0207] In some embodiments, a patient's lymphoproliferative disease, such as leukemia or lymphoma, can be monitored by generating a clonotype profile based on samples (or tissue samples) obtained serially from the patient. The clonotype profile can be generated as described above. In some embodiments, such monitoring may be performed by the following steps: (a) obtaining a sample comprising T cells and / or B cells and / or cell-free DNA from an individual; (b) attaching sequence tags to recombinant nucleic acid molecules of T cell receptor genes or immunoglobulin genes from the sample to form tag-nucleic acid conjugates, wherein at least one recombinant nucleic acid or a copy thereof is attached with a different sequence tag; (c) amplifying the tag-nucleic acid conjugates; (d) sequencing the sample of the tag-nucleic acid conjugates to provide sequence reads, each of which has an error rate and each of which comprises a tag sequence and a recombinant nucleic acid sequence; (e) aligning sequence reads having similar tag sequences to form a group of sequence reads having the same sequence tag; (f) merging the sequence reads in the group to determine clonotypes, wherein whenever the group of sequence reads has at least a 95% probability of being different, the group of sequence reads is merged into a different recombinant nucleic acid sequence; (g) determining a clonotype profile of the sample by determining the level of the clonotype; and (h) determining the level of correlating clonotypes in the clonotype profile. In some embodiments, steps (a) to (h) may be repeated during monitoring of the patient to determine whether the level of the correlating clonotypes demonstrates recurrence of the disease. In some embodiments, the steps of attaching and amplifying may include the following steps: (a) mixing a first set of primers with a recombinant nucleic acid sample from an immune cell expressing an immune receptor and / or cell-free DNA under primer extension conditions in a reaction mixture, wherein each primer of the first set has a receptor-specific portion such that the receptor-specific portion anneals to a different recombinant nucleic acid at a predetermined position and then extends to form a first extension product, and wherein each primer in the first set has a 5' non-complementary end comprising a first primer binding site; (b) removing the unextended primers of the first set from the reaction mixture; and (c) adding the unextended primers of the first set to the sample under primer extension conditions. A second set of primers is added to the reaction mixture, wherein each primer in the second set has a receptor-specific portion such that the receptor-specific portion anneals with the first extension product at a predetermined position and has a 5' non-complementary end comprising a second primer binding site, the primers of the first set and / or the primers of the second set respectively comprise a sequence tag located between the receptor-specific portion and the first or second primer binding site, and wherein each primer of the second set is extended to form a second extension product such that each second extension product comprises the first primer binding site, the second primer binding site, at least one sequence tag, and a recombinant nucleic acid encoding a portion of an immune cell receptor chain. In some embodiments, the step of merging the recombinant nucleic acids comprises: merging such sequence reads whenever there is at least a 99% probability that the sequence reads of different recombinant nucleic acids are different; and in other embodiments, when there is at least a 99.9% probability that they are different.

[0208] The methods of the present invention are also suitable for monitoring minimal residual lesions of cancer in patients, including non-lymphoid cancer or non-bone marrow cancer, wherein the minimal residual lesions have, for example, mutation identification patterns in selected cancer gene groups. Such mutation patterns, that is, the presence, absence and / or level of genes containing such mutations, may indicate the possibility of disease recurrence. In some embodiments, the target polynucleotides used for such monitoring may be exons, partial exons, selected introns and / or gene expression control regions, such as promoters of various genes (referred to herein as "oncogene molecules"). Oncogene molecules can be isolated from tissue samples using conventional techniques, such as exon capture technology, such as TruSeq TM Exome Enrichment Kit (Illumina, San Diego, CA); Frampton et al, Nature Biotechnology, 31(11):1023-1031(2013); etc. After obtaining such oncogene molecules, sequence tags are attached to form tag-nucleic acid conjugates, which are amplified and sequenced according to the present invention.

[0209] Recent cancer genome sequencing studies have shown that there is a significant inconsistency in mutation patterns between different cancers, between different patients with the same cancer, between cells of the same tumor, and between cells of different metastatic sites in the same patient; however, within the same patient, heterogeneous cancer cells often evolve from common progenitor cells, so they share mutations and the evolutionary relationships between cancer cells can be understood over time through a series of measurements, such as Vogelstein et al, Science, 339: 1546-1558 (2013); Ding et al, Nature, 481(7382): 506-510 (2012); etc.; therefore, the mutation patterns associated with a cancer measured in a diagnostic sample provide a means to detect recurrence of the same cancer or a clonal evolutionary version of the cancer.

[0210] Oncogene molecules can be selected from a variety of genes, including but not limited to the genes in Table 1.

[0211] Table I

[0212] Exemplary Oncogenes

[0213]

[0214] In some embodiments, the above-mentioned method for monitoring minimal residual lesions of cancer may include the following steps: (a) obtaining a tissue sample from an individual; (b) attaching a sequence tag to each of a plurality of oncogene molecules in the sample to form a tag-nucleic acid conjugate, wherein at least one nucleic acid or a copy thereof is attached with a different sequence tag, and wherein the oncogene molecule is unique to the individual's cancer; (c) amplifying the tag-nucleic acid conjugate; (d) sequencing the sample of the tag-nucleic acid conjugate to provide a sequence read having an error rate and comprising a tag sequence and an oncogene sequence; (e) aligning sequence reads with similar tag sequences to form a sequence read group with the same sequence tag; (f) merging the oncogene sequences in the group to determine the sequence of the oncogene molecule, wherein whenever the oncogene sequence group has at least a 95% probability of being different, the sequence read group is merged into a different oncogene molecule; and (g) detecting the presence, absence and / or level of oncogene molecules unique to the individual's cancer in the oncogene molecular spectrum. In some embodiments, the step of merging oncogene sequences comprises merging sequence reads of different oncogene molecules whenever there is at least a 99% probability that the sequence reads are different; and in other embodiments there is at least a 99.9% probability that the sequence reads are different.

[0215] Detecting carryover contamination using sequence tags

[0216] Carryover contamination is a significant problem for techniques including nucleic acid amplification, e.g., Borst et al, Eur. J. Clin. Microbiol. Infect. Dis., 23(4): 289-299 (2004); Aslanzadeh, Ann. Clin. Lab. Sci., 34(4): 389-396 (2004); etc. Such contamination occurs when trace amounts of sample foreign nucleic acid are inadvertently amplified when the sample is analyzed and affect the measurement results. In worse cases, carryover contamination in a patient's medical sample can lead to a false positive interpretation of the analysis result. Foreign nucleic acid can come from a source unrelated to the specific patient; for example, it can come from a sample from another patient. Alternatively, the exogenous nucleic acid may be from a source related to the patient; for example, it may be from a different sample from the same patient processed in the past in the same laboratory or from an analytical reaction on a different sample from the same patient processed in the past in the same laboratory.

[0217] Carryover contamination in a clinical setting is particularly challenging when measuring highly complex populations of related nucleic acids, such as recombinant nucleic acid populations encoding immune molecules such as T cell receptors or immunoglobulins. The challenge exists because it is difficult to determine whether a sequence read or clonotype is part of the true diversity of the target sample or whether the sequence read or clonotype is derived from an exogenous nucleic acid, such as a sample of another patient or a previous sample of the same patient processed in the same type of analysis in the same laboratory. In one aspect of the present invention, such carryover contamination can be detected using sequence tags, so that not only can the clonotype be determined from the sequence read, but also whether the sequence tag is derived from the current sample or from another sample. This is achieved by the following steps: saving a record of the sequence tags determined from each patient sample, and then comparing the sequence tags of the current measurement with the sequence tags of the previous measurement whenever a subsequent measurement is performed. These sequence tag records associated with the clonotype are conveniently saved as electronic records on a large-capacity storage device, because there are a large number of tags from each measurement, and this facilitates searching the electronic records and comparing the electronic records using conventional algorithms. If there is a match, then the most likely cause is carryover contamination, provided that the population of sequence tags used in the measurement is large enough. The same exemplary ratio of the size of the clonal population to the sequence tag population used for sampling labeling described above is applicable to detecting carryover contamination. In one embodiment, this ratio is 100:1 or greater.

[0218] A variety of search methods or algorithms can be used to compare the measured clonotypes with the database clonotypes. Many conventional sequence alignment and search algorithms are publicly available and have been described in the following references incorporated by reference: Mount, Bioinformatics Sequence and Genome Analysis, Second Edition (Cold Spring Harbor Press, 2004); Batzoglou, Briefings in Bioinformatics, 6: 6-22 (2005); Altschul et al, J. Mol. Biol., 215 (3): 403-410 (1990); Needleman and Wunsch, J. Mol. Biol., 48: 443-453 (1970); Smith and Waterman, Advances in Applied Mathematics, 2: 482-489 (1981); etc.

[0219] In some embodiments, the above-mentioned method for detecting and determining contamination (such as carryover contamination) in a sample caused by substances originating from different samples may include the following steps: (a) obtaining a tissue sample from an individual; (b) attaching a sequence tag to an oncogene molecule or a recombinant nucleic acid to form a tag-nucleic acid conjugate, wherein at least one nucleic acid or a copy thereof is attached with a different sequence tag, and wherein the oncogene molecule is unique to the cancer of the individual; (c) amplifying the tag-nucleic acid conjugate; (d) sequencing the sample of the tag-nucleic acid conjugate to provide sequence reads, each of which has an error rate and each contains a tag sequence and an oncogene sequence or a recombinant nucleic acid sequence; (e) comparing the tag sequence with a tag sequence from other tissue samples determined separately; and (f) determining whether there is contamination and / or the degree of contamination by the consistency of one or more tag sequences with any tag sequence from other tissue samples determined separately. Once the tag sequences are determined in the analysis, these tag sequences can be compared with tag sequences in a tag sequence database recorded based on the analysis of other patients. Such a comparison step can be performed at the time of analysis, or such a step can be implemented retrospectively, for example, at a certain period of time after the analysis. In one embodiment, the sequence tags are attached to recombinant nucleic acids in a tissue sample (such as blood or bone marrow) from an individual suffering from a lymphoproliferative disease such as lymphoma. In another embodiment, the sequence tags are attached to an oncogene molecule, as described above.

[0220] In other embodiments in which recombinant nucleic acids are monitored to prevent cross-contamination of tissue samples, the following attachment and amplification steps can be performed: (a) in a reaction mixture, a first set of primers is mixed with a recombinant nucleic acid sample from T cells and / or cell-free DNA under primer extension conditions, wherein each primer of the first set has a receptor-specific portion such that the receptor-specific portion anneals with a different recombinant nucleic acid at a predetermined position and then extends to form a first extension product, and wherein each primer in the first set has a 5' non-complementary end comprising a first primer binding site; (b) unextended primers of the first set are removed from the reaction mixture; (c) a second set of primers is added to the reaction mixture under primer extension conditions, wherein each primer in the second set has a receptor-specific portion; portion, so that the receptor-specific portion anneals and pairs with the first extension product at a predetermined position and has a 5' non-complementary end containing a second primer binding site, the first group of primers and / or the second group of primers respectively contain a sequence tag located between the receptor-specific portion and the first or second primer binding site, and wherein each primer of the second group is extended to form a second extension product, so that each second extension product contains the first primer binding site, the second primer binding site, at least one sequence tag and a recombinant nucleic acid encoding a portion of the immune receptor chain; and (d) performing a polymerase chain reaction in a reaction mixture to form an amplicon, wherein the polymerase chain reaction uses a forward primer specific for the first primer binding site and a reverse primer specific for the second primer binding site.

[0221] Reagent test kit

[0222] The present invention includes a variety of kits for performing the methods of the present invention. In some embodiments, the kit includes (a) a set of forward primers and a set of reverse primers, which are used to amplify recombinant nucleic acids encoding multiple immune receptor chains in multiplex PCR, wherein the forward primers and / or reverse primers each have a target-specific portion, a sequence tag, and a common primer binding site; and (b) a primer removal element for removing primers that are not incorporated into the primer set after at least the first extension (i.e., primers that are not extended). In some embodiments, the kit also includes universal primers specific to a common primer binding site. In some embodiments, the kit also includes written instructions for using kit components in the methods of the present invention. In some embodiments, the kit also includes forward and reverse primers specifically for amplifying recombinant nucleic acids encoding IgH (VDJ), IgH (DJ) and IgK. In some embodiments, the kit also includes forward and reverse primers specifically for amplifying recombinant nucleic acids encoding TCRβ, TCRδ and TCRγ. In some embodiments, the kit also includes an internal standard, which includes multiple nucleic acids having a representative length and composition of a target recombinant nucleic acid, wherein the internal standard is provided at a known concentration. In some embodiments, the kit includes a single-stranded exonuclease as a primer removal element, such as E. coli exonuclease I. In some embodiments, the kit includes a spin column capable of size sorting double-stranded DNA as a primer removal element.

[0223] Although the present invention has been described in conjunction with several specific exemplary embodiments, those skilled in the art will recognize that various changes may be made to the present invention without departing from the spirit and scope of the present invention. In addition to those discussed above, the present invention may also be applied to various sensor implementations and other topics.

[0224] definition

[0225] Unless otherwise specifically defined herein, the terms and symbols of nucleic acid chemistry, biochemistry, genetics, and molecular biology used herein follow those in standard treatises and textbooks in the field, such as Kornberg and Baker, DNA Replication, Second Edition (WH Freeman, New York, 1992); Lehninger, Biochemistry, Second Edition (Worth Publishers, New York, 1975); Strachan and Read, Human Molecular Genetics, Second Edition (Wiley-Liss, New York, 1999); Abbas et al, Cellular and Molecular Immunology, 6 th edition (Saunders, 2007) (Abbas et al., Cellular and Molecular Immunology, 6th edition (Saunders Publishing, 2007)).

[0226] "Alignment" refers to a method of comparing a test sequence (such as a sequence read) to one or more reference sequences based on a certain sequence distance measure to determine which reference sequence or which portion of a reference sequence is closest. An exemplary method for aligning nucleotide sequences is the Smith-Waterman algorithm. Distance measures may include Hamming distance, Levenshtein distance, and the like. Distance measures may include a portion related to the quality values ​​of the nucleotides of the compared sequences.

[0227] "Amplicon" refers to the product of a polynucleotide amplification reaction, i.e., a clonal population of polynucleotides, which may be single-stranded or double-stranded, and which is replicated from one or more starting sequences. The one or more starting sequences may be one or more copies of the same sequence, or they may be a mixture of different sequences. Amplicon can be produced by a variety of amplification reactions, the product of which includes replicas of one or more starting nucleic acids or target nucleic acids. In one aspect, the amplification reaction that produces an amplicon is "template-driven" because the base pairing of the reactants (whether nucleotides or oligonucleotides) has complements in the template polynucleotides required to produce the reaction product. In one aspect, the template-driven reaction is a primer extension using a nucleic acid polymerase or an oligonucleotide connection using a nucleic acid ligase. Such reactions include, but are not limited to, polymerase chain reaction (PCR), linear polymerase reaction, nucleic acid sequence-based amplification (NASBA), rolling circle amplification, etc., as disclosed in the following references incorporated herein by reference: Mullis et al., U.S. Pat. Nos. 4,683,195, 4,965,188, 4,683,202, 4,800,159 (PCR); Gelfand et al., U.S. Pat. No. 5,210,015 (real-time PCR using "taqman" probes); Wittwer et al., U.S. Pat. No. 6,174,670; Kacian et al., U.S. Pat. No. 5,399,491 ("NASBA"); Lizardi, U.S. Pat. No. 5,854,033; Aono et al., Japanese Patent Publication No. JP 4-262799 (rolling circle amplification); etc. In one aspect, the amplicon of the present invention is generated by PCR. If a detection reaction is provided that can measure the reaction products as the amplification reaction proceeds, the amplification reaction can be a "real-time" amplification, for example, "real-time PCR" as described below or "real-time NASBA" as described in Leone et al, Nucleic Acids Research, 26:2150-2155 (1998) and similar references. As used herein, the term "amplification" refers to performing an amplification reaction. A "reaction mixture" refers to a solution containing all the necessary reactants for performing a reaction, which may include, but is not limited to, buffers, salts, cofactors, scavengers, etc., to maintain the pH at a selected level during the reaction.

[0228] As used herein, "clonality" refers to a measure of the degree to which the distribution of clonotype abundances among clonotypes of a lineage is skewed to a single or a few clonotypes. In general, clonality is the opposite measure of clonotype diversity. Many metrics or statistical data are available from ecology that describe species-abundance relationships that can be used for clonality metrics according to the present invention, such as Chapters 17 and 18 of Pielou, An Introduction to Mathematical Ecology, (Wiley-Interscience, 1969). In one aspect, the clonality metric used by the present invention is a function of the clonotype spectrum (i.e., the number of different clonotypes detected and their abundance), such that after the clonotype spectrum is determined, clonality can be calculated from the clonotype spectrum to obtain a single number. One clonality metric is the Simpson metric, which is simply the probability that two randomly drawn clonotypes will be the same. Other clonality metrics include the information-based metric disclosed in Pielou (cited above) and the McIntosh diversity index.

[0229] "Clonotype" refers to a recombinant nucleic acid of a lymphocyte encoding an immune receptor or a portion thereof. More specifically, "clonotype" refers to a recombinant nucleic acid that is usually extracted from a T cell or B cell but can also come from a cell-free source, which encodes a T cell receptor (TCR) or a B cell receptor (BCR) or a portion thereof. In various embodiments, the clonotype may encode all or part of the VDJ rearrangement of IgH, the DJ rearrangement of IgH, the VJ rearrangement of IgK, the VJ rearrangement of IgL, the VDJ rearrangement of TCRβ, the DJ rearrangement of TCRβ, the VJ rearrangement of TCRα, the VJ rearrangement of TCRγ, the VDJ rearrangement of TCRδ, the VD rearrangement of TCRδ, the Kde-V rearrangement, etc. The clonotype may also encode a translocation breakpoint region involving an immune receptor gene (such as Bcl1-IgH or Bcl1-IgH). In one aspect, the clonotype has a sufficiently long sequence to characterize or reflect the diversity of the immune molecule from which it is derived; therefore, the length of the clonotype may vary greatly. In some embodiments, the clonotype is in the range of 25 to 400 nucleotides in length; in other embodiments, the clonotype is in the range of 25 to 200 nucleotides in length.

[0230] A "clonotype profile" refers to a list of different clonotypes derived from a lymphocyte population and their relative abundance, where, for example, the relative abundance can be expressed as a frequency in a given population (that is, based on a numerical value between 0 and 1). Typically, a lymphocyte population is obtained from a tissue sample. The term "clonotype profile" relates to the immunological concept of immune "lineage" as described in, for example, the following references, but is broader than this concept: Arstila et al, Science, 286:958-961 (1999); Yassai et al, Immunogenetics, 61:493-502 (2009); Kedzierska et al, Mol. Immunol., 45(3):607-618 (2008); etc. The term "clonotype profile" includes a large list and abundance of rearranged nucleic acids encoding immune receptors, which may be derived from a selected subset of lymphocytes (e.g., tissue-infiltrating lymphocytes, a subset of immune phenotypes, etc.), or which may encode portions of immune receptors with lower diversity than the complete immune receptors. In some embodiments, the clonotype profile may include at least 10 3 In other embodiments, the clonotype profile may include at least 10 4 In other embodiments, the clonotype profile may include at least 10 5 In other embodiments, the clonotype profile may include at least 10 6In one aspect, the clonotype spectrum is a set of different recombinant nucleotide sequences (and their abundance) encoding T cell receptors (TCR) or B cell receptors (BCR) or fragments thereof in a lymphocyte population of an individual, wherein the nucleotide sequences in the set have a one-to-one correspondence with lymphocytes or their clonal subpopulations that are different from substantially all lymphocytes in the population. In one aspect, the nucleic acid segments defining the clonotypes are selected so that their diversity (i.e., the number of different nucleic acid sequences in the set) is large enough that substantially every T cell or B cell or their clones in the individual carry a unique nucleic acid sequence of this lineage. That is, each different clone of the sample preferably has a different clonotype. In other aspects of the invention, the lymphocyte population corresponding to the lineage may be circulating B cells, or may be circulating T cells, or may be a subpopulation of any of the above populations, including but not limited to CD4+T cells or CD8+T cells, or other subpopulations defined by cell surface markers, etc. Such subpopulations can be obtained by collecting samples from specific tissues (e.g., bone marrow or lymph nodes, etc.), or by sorting or enriching cells from samples (e.g., peripheral blood) based on one or more cell surface markers, size, morphology, etc. In yet other aspects, the lymphocyte population corresponding to the lineage can be derived from disease tissue (e.g., tumor tissue, infected tissue, etc.). In one embodiment, the clonal type spectrum comprising human TCR β chain or fragment thereof is included in 0.1×10 6 to 1.8×10 6 In the range of 0.5×10 6 to 1.5×10 6 In the range of 0.8×10 6 to 1.2×10 6 In another embodiment, the clonal profile comprising human IgH chains or fragments thereof comprises a plurality of different nucleotide sequences within a range of 0.1×10 6 to 1.8×10 6 In the range of 0.5×10 6 to 1.5×10 6 In the range of 0.8×10 6to 1.2×106. In a specific embodiment, the clonotype spectrum of the present invention comprises a set of nucleotide sequences encoding substantially all segments of the V(D)J region of the IgH chain. In one aspect, "substantially all" as used herein means that each segment has a relative abundance of 0.001% or more; or on the other hand, "substantially all" as used herein means that each segment has a relative abundance of 0.0001% or more. In another specific embodiment, the clonotype spectrum of the present invention comprises a set of nucleotide sequences encoding substantially all segments of the V(D)J region of the TCRβ chain. In another embodiment, the clonotype spectrum of the present invention comprises a set of nucleotide sequences of segments of the V, D, and J regions of the TCRβ chain in the range of 25 to 200 nucleotides in length. In another embodiment, the clonotype spectrum of the present invention comprises a set of nucleotide sequences of segments of the V, D, and J regions of the IgH chain in the range of 25 to 200 nucleotides in length. In another embodiment, the clonotype spectrum of the present invention comprises a number of multiple different nucleotide sequences substantially equivalent to the number of lymphocytes expressing different IgH chains. In another embodiment, the clonotype profile of the present invention comprises a plurality of different nucleotide sequences in a number substantially equivalent to the number of lymphocytes expressing different TCRβ chains. In yet another embodiment, "substantially equivalent" means that the clonotype profile will have a 99% probability of comprising such a nucleotide sequence: the nucleotide sequence encodes an IgH or TCRβ or a portion thereof carried or expressed at a frequency of 0.001% or greater by each of the lymphocyte population of the individual. In yet another embodiment, "substantially equivalent" means that the nucleotide sequence repertoire will have a 99% probability of comprising such a nucleotide sequence: the nucleotide sequence encodes an IgH or TCRβ or a portion thereof carried or expressed at a frequency of 0.0001% or greater by each lymphocyte. In some embodiments, the clonotype profile is obtained from a sample comprising 105 to 107 lymphocytes. Such a number of lymphocytes can be obtained from a 1-10 mL peripheral blood sample.

[0231] "Complementarity determining region" (CDR) refers to the region of an immunoglobulin (i.e., antibody) or T-cell receptor in which the conformation of the molecule and the antigen are complementary, thereby determining the molecule's specificity for and contact with the specific antigen. T-cell receptors and immunoglobulins each have three CDRs: CDR1 and CDR2 are present in the variable (V) domain, while CDR3 comprises part of the V domain, the entire multivariate domain (D) (heavy chain only) and the joining domain (J), and part of the constant domain (C).

[0232] A "clonotype database" refers to a collection of clonotypes formatted and arranged to simplify and expedite searching, comparison, and retrieval. In some embodiments, the clonotype database comprises a collection of clonotypes encoding the same region or segment of an immune receptor. In some embodiments, the clonotype database comprises clonotypes from clonotype profiles of multiple individuals. In some embodiments, the clonotype database comprises at least 10 clonotype profiles from at least 10 individuals. 4 In some embodiments, the clonotype database comprises at least 10 6 clonotypes, or at least 10 8 clonotypes, or at least 10 9 clonotypes, or at least 10 10 The clonotype database may be a public database containing clonotypes, such as the IMGT database (www.imgt.org), as described in Nucleic Acids Research, 31:307-310 (2003). The clonotype database may be in FASTA format, and the BLAST algorithm (such as Altschul et al, J. Mol. Biol., 215 (3): 403-410 (1990) or similar algorithms may be used to search or compare clonotype database entries.

[0233] "Merging" refers to treating two candidate clonotypes with sequence differences as the same clonotype by determining that such sequence differences are due to experimental or measurement error rather than true biological differences. In one aspect, the sequence of the more frequent candidate clonotype is compared to the sequence of the less frequent candidate clonotype, and if a predetermined criterion is met, the number of the less frequent candidate clonotype is added to the number of the more frequent clonotype, and the less frequent candidate clonotype is then ignored. That is, the read counts associated with the less frequent candidate clonotype are added to the read counts of the more frequent candidate clonotype, and the more frequent candidate clonotype and the less frequent candidate clonotype are treated as the same clonotype; that is, the observed differences between the two candidate clonotypes are determined to be due to error (e.g., sequencing error, amplification error, etc.). In some embodiments, the predetermined criterion is a likelihood function that depends on factors such as the relative frequencies of the candidate clonotypes being compared, the number of positions where the candidate clonotypes differ, the quality scores of the positions, etc.

[0234] "Complementarity determining region" (CDR) refers to the region of an immunoglobulin (i.e., antibody) or T-cell receptor in which the conformation of the molecule and the antigen are complementary, thereby determining the molecule's specificity for and contact with the specific antigen. T-cell receptors and immunoglobulins each have three CDRs: CDR1 and CDR2 are present in the variable (V) domain, while CDR3 comprises part of the V domain, the entire multivariate domain (D) (heavy chain only) and the joining domain (J), and part of the constant domain (C).

[0235] As used herein, "contamination" refers to the presence of nucleic acid from one individual in a tissue sample from another individual. In one aspect, "contamination" refers to the presence of nucleic acid that did not originate from the patient, which can affect the interpretation of the clonotype profile of the patient.

[0236] "Genetic identification" refers to the unique correspondence between an individual and a set of values ​​(or states) of genetic markers from one or more loci for that individual.

[0237] "Genetic marker" refers to a polymorphic segment of DNA at a locus that can be used to identify an individual. A genetic marker can be identified by its sequence or by its adjacent or flanking sequence. Typically, a genetic marker may have multiple sequences or values ​​in different individuals of a population. Exemplary genetic markers include, but are not limited to, short tandem repeats (STRs), single nucleotide polymorphisms (SNPs), and the like. The polymorphic segment of DNA can be genomic DNA or reverse transcribed RNA. In one embodiment, the polymorphic segment is genomic DNA. In one embodiment, the genetic markers used in the present invention are identified by amplification and sequencing using conventional techniques. In another embodiment, the genetic markers are amplified and sequenced together with immune molecules in the process of generating a clonal spectrum.

[0238] "Internal standard" refers to a nucleic acid sequence that is treated in the same manner as one or more target polynucleotides in the same reaction, and is used to perform absolute or relative quantification of the target polynucleotides in the sample. In one aspect, the reaction is an amplification reaction, such as PCR. The internal standard can be endogenous or exogenous. That is, the internal standard can be naturally present in the sample, or it can be added to the sample before the reaction. In one aspect, one or more exogenous internal standard sequences can be added to the reaction mixture at a predetermined concentration to provide a calibration sample, and the amplified sequence can be compared with the calibration sample to determine the amount of its corresponding target polynucleotide in the sample. For those of ordinary skill in the art, the selection of the number, sequence, length and other characteristics of the exogenous internal standard belongs to conventional design selection. The endogenous internal standard is also referred to as a "reference sequence" in this article, which is a sequence naturally present in the sample, which corresponds to the lowest regulated gene that exhibits a constant and cell cycle-independent transcription level, such as Selvey et al, Mol. Cell Probes, 15: 307-311 (2001) (Selvey et al., "Molecular Cell Probes", Vol. 15, p. 307-311, 2001). Exemplary internal standards include, but are not limited to, sequences from the following genes: GAPDH, β2-microglobulin, 18S ribosomal RNA, and β-actin.

[0239] "Kit" refers to any delivery system for delivering materials or reagents for implementing the inventive method. In the context of the inventive method, such delivery systems include systems that allow reaction reagents (e.g., primers, enzymes, internal standards, etc. placed in appropriate containers) and / or support materials (e.g., buffers, written instructions on how to analyze, etc.) to be stored, transported, or delivered from one location to another. For example, a kit includes one or more packages (e.g., boxes) containing relevant reaction reagents and / or support materials. These contents can be delivered to a target receptor together or individually. For example, a first container may contain an enzyme for analysis, and a second container contains a primer.

[0240] "Minimal residual disease" refers to cancer cells that remain after treatment. The term is often used in conjunction with the treatment of lymphoma and leukemia.

[0241] "Lymphoid or myeloproliferative disease" refers to any abnormal proliferative disease in which one or more nucleotide sequences encoding one or more rearranged immune receptors can be used as a marker for monitoring such disease. "Lymphoid or myeloid neoplasm" refers to an abnormal proliferation of lymphocytes or myeloid cells, whether malignant or non-malignant. Lymphoma is a malignant lymphoma. Myeloma is a malignant myeloid neoplasm. Lymphoid tumors and myeloid tumors are caused by or are associated with lymphoid or myeloproliferative diseases, and include, but are not limited to, follicular lymphoma, chronic lymphocytic leukemia (CLL), acute lymphocytic leukemia (ALL), chronic myeloid leukemia (CML), acute myeloid leukemia (AML), Hodgkin's lymphoma and non-Hodgkin's lymphoma, multiple myeloma (MM), monoclonal gammopathy of undetermined significance (MGUS), mantle cell lymphoma (MCL), diffuse large B-cell lymphoma (DLBCL), myelodysplastic syndrome (MDS), T-cell lymphoma, etc., for example, Jaffe et al, Blood, 112:4384-4399 (2008); Swerdlow et al, WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues (e.g., Jaffe et al, Blood, 112:4384-4399 (2008); Swerdlow et al, WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues (e.g., Jaffe et al, Blood, 112:4384-4399 (2008); Swerdlow et al, WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues (e.g., th )(IARC Press, 2008)(Swerdlow et al., WHO Classification of Tumors of Hematopoietic and Lymphoid Tissue (4th Edition) (International Agency for Research on Cancer (IARC) Press Release, 2008)). As used herein, "B-cell cancer" refers to a lymphoid or myeloid neoplasm involving a B cell or a cell that develops from a B cell (e.g., a plasma cell). Similarly, "T-cell cancer" refers to a lymphoid or myeloid neoplasm involving a T cell or a cell that develops from a T cell.

[0242] "Percent homology," "percent identity," or similar terms used in reference to a comparison of a reference sequence and another sequence ("comparison sequence") means that in an optimal alignment between the two sequences, the comparison sequence is identical to the reference sequence at a number of subunit positions corresponding to the indicated percentage, where the subunits are nucleotides for polynucleotide comparisons or amino acids for polypeptide comparisons. As used herein, an "optimal alignment" of the compared sequences is one that maximizes matches between subunits and minimizes the number of gaps used in constructing the alignment. Percent identity can be determined using commercially available algorithm implementations such as the algorithm described by Needleman and Wunsch, J. Mol. Biol., 48:443-453 (1970) ("GAP" program of Wisconsin Sequence Analysis Package, Genetics Computer Group, Madison, WI) or similar algorithms. Other software packages used in the art for constructing alignments and calculating percent identity or other similarity metrics include the "BestFit" program (Wisconsin Sequence Analysis Package, Genetics Computer Group, Madison, WI), based on the algorithm of Smith and Waterman, Advances in Applied Mathematics, 2:482-489 (1981). In other words, for example, to obtain a polynucleotide having a nucleotide sequence that is at least 95% identical to a reference nucleotide sequence, up to 5% of the nucleotides in the reference sequence may be deleted or substituted for another nucleotide, or up to 5% of the total number of nucleotides in the reference sequence may be inserted into the reference sequence.

[0243] "Polymerase chain reaction" or "PCR" refers to a reaction for amplifying a specific DNA sequence in vitro by simultaneous primer extension of complementary DNA strands. In other words, PCR is a reaction for preparing multiple copies or replicas of a target nucleic acid flanked by a primer binding site, and this reaction includes one or more repetitions of the following steps: (i) denaturing the target nucleic acid, (ii) annealing and pairing the primer to the primer binding site, and (iii) extending the primer using a nucleic acid polymerase in the presence of nucleoside triphosphates. As used herein, the terms "forward primer" and "upstream primer" are used interchangeably, and the terms "reverse primer" and "downstream primer" are used interchangeably. Also as used herein, if the double-stranded target polynucleotide exhibits a 5'→3' orientation from left to right for its sense strand, the forward primer will bind to the antisense strand on the left and extend to the right, and the reverse primer will bind to the sense strand on the right and extend to the left. Typically, the reaction is cycled using different temperatures optimized for each step in a thermal cycler. The specific temperature, duration of each step, and rate of change between steps depend on many factors well known to those of ordinary skill in the art, as exemplified by the following references: McPherson et al, editors, PCR: A Practical Approach and PCR2: A Practical Approach (IRL Press, Oxford, 1991 and 1995, respectively). For example, in conventional PCR using Taq DNA polymerase, a double-stranded target nucleic acid may be denatured at a temperature >90°C, primers annealed at a temperature in the range of 50-75°C, and primers extended at a temperature in the range of 72-78°C. The term "PCR" encompasses derivative forms of the reaction, including but not limited to RT-PCR, real-time PCR, nested PCR, quantitative PCR, multiplex PCR, etc. The reaction volume ranges from a few hundred nanoliters (e.g., 200 nL) to a few hundred μL (e.g., 200 μL). "Reverse transcription PCR" or "RT-PCR" refers to PCR in which the target RNA is first converted into a complementary single-stranded DNA by reverse transcription, which is then amplified, as described, for example, in Tecott et al., U.S. Pat. No. 5,168,038, which is incorporated herein by reference. "Real-time PCR" refers to PCR in which the amount of the reaction product, i.e., the amplicon, is monitored as the reaction proceeds.There are many forms of real-time PCR, which differ primarily in the detection chemistry used to monitor the reaction products, such as Gelfand et al., U.S. Pat. No. 5,210,015 ("taqman"); Wittwer et al., U.S. Pat. Nos. 6,174,670 and 6,569,627 (intercalating dyes); Tyagi et al., U.S. Pat. No. 5,925,517 (molecular beacons); these patents are incorporated herein by reference. Detection chemistry for real-time PCR is described in Mackay et al, Nucleic Acids Research, 30:1292-1305 (2002), which is also incorporated herein by reference. "Nested PCR" refers to a two-stage PCR in which the amplicon of the first PCR becomes the sample for the second PCR, which uses a new set of primers, at least one of which binds to an internal position of the first amplicon. As used herein, "initial primers" for nested amplification reactions refer to primers used to generate the first amplicon, and "second primers" refer to one or more primers used to generate the second or nested amplicon. "Multiplex PCR" refers to a PCR in which multiple target sequences (or a single target sequence and one or more reference sequences) are simultaneously performed in the same reaction mixture, such as Bernard et al, Anal. Biochem., 273: 221-228 (1999) (Bernard et al., "Analytical Biochemistry", Vol. 273, pp. 221-228, 1999) (dual-color real-time PCR). Typically, a different primer set is used for each sequence amplified. "Quantitative PCR" refers to a PCR designed to measure the abundance of one or more specific target sequences in a sample or specimen. Quantitative PCR includes both absolute quantification of such target sequences and relative quantification thereof. Quantitative measurements are performed using one or more reference sequences or internal standards, which can be analyzed individually or together with the target sequence. The reference sequence may be endogenous or exogenous to the sample or specimen, and in the latter case may comprise one or more competing templates. Typical endogenous reference sequences include segments of transcripts of the following genes: β-actin, GAPDH, β2-microglobulin, ribosomal RNA, and the like.Techniques for quantitative PCR are well known to those of ordinary skill in the art, as exemplified in the following references incorporated by reference: Freeman et al, Biotechniques, 26: 112-126 (1999); Becker-Andre et al, Nucleic Acids Research, 17: 9437-9447 (1989); Zimmerman et al, Biotechniques, 21: 268-279 (1996); Diviacco et al, Biotechniques, 21: 268-279 (1996); al, Gene, 122: 3013-3020 (1992); Becker-Andre et al, Nucleic Acids Research, 17: 9437-9446 (1989); etc.

[0244] "Primer" refers to a natural or synthetic oligonucleotide that, after forming a duplex with a polynucleotide template, can serve as a starting point for nucleic acid synthesis and extend from its 3' end along the template to form an extended duplex. Extension of the primer is usually performed using a nucleic acid polymerase such as a DNA or RNA polymerase. The sequence of nucleotides added during the extension process is determined by the sequence of the template polynucleotide. Typically, a DNA polymerase is used to extend the primer. The length of the primer is usually in the range of 14 to 40 nucleotides, or in the range of 18 to 36 nucleotides. Primers are used in a variety of nucleic acid amplification reactions, for example, linear amplification reactions using a single primer, or polymerase chain reactions using two or more primers. Guidance on the length and sequence of primers selected for specific applications is well known to those of ordinary skill in the art, as confirmed by the following references incorporated by reference: Dieffenbach, editor, PCR Primer: A Laboratory Manual, 2 nd Edition (Cold Spring Harbor Press, New York, 2003) (Dieffenbach, ed., PCR Primers: A Laboratory Manual, 2nd Edition (Cold Spring Harbor Press, New York, 2003)).

[0245] "Quality score" refers to a measure of the probability of correctly assigning a base to a particular sequence position. For specific cases, such as, a variety of methods for calculating quality scores for bases read by different sequencing chemistries, detection systems, base calling algorithms, etc. are well known to those of ordinary skill in the art. In general, the quality score value is monotonically related to the probability of a correct base call. For example, a quality score or Q of 10 may indicate a 90% probability that the base is correctly called, a Q of 20 may indicate a 99% probability that the base is correctly called, and so on. For some sequencing platforms, especially those using sequencing-by-synthesis chemistries, the average quality score decreases as a function of sequence read length, such that the quality score at the beginning of the sequence read is higher than the quality score at the end of the sequence read, and this decrease is caused by phenomena such as incomplete extension, carry forward extension, template loss, polymerase loss, capping failure, deprotection failure, etc.

[0246] "Sequence read" refers to a nucleotide sequence determined by a sequence or data stream generated by a sequencing technology, such determination being achieved, for example, by means of base calling software associated with the technology (such as base calling software from a DNA sequencing platform supplier). A sequence read typically includes a quality score for each nucleotide in the sequence. Typically, a sequence read is obtained by extending a primer along a template nucleic acid, for example, with a DNA polymerase or a DNA ligase. Data is generated by recording signals associated with such extensions, such as optical signals, chemical signals (e.g., pH changes), or electrical signals. This initial data is converted into a sequence read.

[0247] A "sequence tag" (or "tag") or "barcode" refers to an oligonucleotide that is attached to a polynucleotide or template molecule and is used to identify and / or track the polynucleotide or template in a reaction or a series of reactions. Each sequence tag has a nucleotide sequence that is sometimes referred to herein as a "tag sequence." The sequence tag can be attached to the 3'-end or 5'-end of a polynucleotide or template, or can be inserted into the interior of such a polynucleotide or template to form a linear or circular conjugate, which is sometimes referred to herein as a "tagged polynucleotide" or "tagged template" or "tag-polynucleotide conjugate," "tag-molecule conjugate," or the like. Sequence tags can vary widely in size and composition; the following references incorporated herein by reference provide guidance on selecting a set of sequence tags suitable for use in a particular embodiment: Brenner, U.S. Pat. No. 5,635,400; Brenner and Macevicz, U.S. Pat. No. 7,537,897; Brenner et al, Proc. Natl. Acad. Sci., 97:1665-1670 (2000); Church et al, European Patent Publication No. 0 303 459; Shoemaker et al, Nature Genetics, 14: 450-456 (1996) (Shoemaker et al., Nature Genetics, Vol. 14, pp. 450-456, 1996); Morris et al., European Patent Publication No. 0799897A1; Wallace, U.S. Pat. No. 5,981,179; etc. The selection of a particular tag length and / or composition may depend on several factors, including, but not limited to: the sequencing technology used to decode the tags; the number of distinguishable tags required to unambiguously identify a group of target polynucleotides; how different a group of tags must be in order to ensure reliable identification, for example, to avoid cross-hybridization or misidentification due to sequencing errors; etc. In some embodiments, the length of the sequence tags may each be in the range of 6 to 100 nucleotides, or in the range of 10 to 100 nucleotides, or in the range of 12 to 50 nucleotides, or in the range of 12 to 25 nucleotides.In some embodiments, a set of sequence tags is used, wherein each sequence tag in a set has a unique nucleotide sequence that differs by at least four bases from the sequences of all other tags in the same set; in other embodiments, a set of sequence tags is used, wherein the sequence of each tag in a set differs by at least five bases from the sequences of all other tags in the same set; in still other embodiments, a set of sequence tags is used, wherein the sequence of each tag in a set differs by at least 10% in its nucleotides from the sequences of all other tags in the same set; or in other embodiments, their nucleotides differ by at least 25%; or in other embodiments, their nucleotides differ by at least 50%.

[0248] Sequence Listing

[0249] <110> Adaptive Biotech

[0250] <120> Large-scale biomolecular analysis using sequence tags

[0251] <130> 848US00(37623-739.601)

[0252] <150> 61 / 841878

[0253] <151> 2013-07-01

[0254] <150> 62 / 001580

[0255] <151> 2014-05-21

[0256] <160> 6

[0257] <170> PatentIn version 3.5

[0258] <210> 1

[0259] <211> twenty four

[0260] <212> DNA

[0261] <213> Artificial sequence

[0262] <220>

[0263] <223> Primers

[0264] <400> 1

[0265] agttctggct aacctgtaga gcca twenty four

[0267] <210> 2

[0268] <211> twenty four

[0269] <212> DNA

[0270] <213> Artificial sequence

[0271] <220>

[0272] <223> Primers

[0273] <400> 2

[0274] agttcgggct aacctgtcga gcca twenty four

[0276] <210> 3

[0277] <211> twenty four

[0278] <212> DNA

[0279] <213> Artificial sequence

[0280] <220>

[0281] <223> Primers

[0282] <400> 3

[0283] agttccggct aacctgtcga gcca twenty four

[0285] <210> 4

[0286] <211> twenty two

[0287] <212> DNA

[0288] <213> Artificial sequence

[0289] <220>

[0290] <223> Primers

[0291] <220>

[0292] <221> misc_feature

[0293] <222> (1)..(22)

[0294] <223> n is a, c, g, or t

[0295] <400> 4

[0296] nnnnnnnnnn nnnnnnnnnn twenty two

[0298] <210> 5

[0299] <211> 12

[0300] <212> DNA

[0301] <213> Artificial sequence

[0302] <220>

[0303] <223> Primers

[0304] <400> 5

[0305] gtattttttt ct 12

[0307] <210> 6

[0308] <211> 13

[0309] <212> DNA

[0310] <213> Artificial sequence

[0311] <220>

[0312] <223> Primers

[0313] <400> 6

[0314] ttcagggggg gct 13

Claims

1. Use of sequence tags in the preparation of reagents for detecting minimal residual lesions in patients treated for cancer, the use comprising: (a) attaching a sequence tag to each of a plurality of recombinant nucleic acids in a sample comprising T cells and / or B cells and / or cell-free DNA or RNA obtained from the patient to form a tag-nucleic acid conjugate, wherein at least one recombinant nucleic acid or a copy thereof has a different sequence tag attached and is specific to the cancer of the patient, and wherein the attaching comprises: (i) mixing a first set of primers with the sample under primer extension conditions in a reaction mixture, wherein each primer in the first set comprises a receptor-specific portion, a 5' non-complementary end comprising a first primer binding site, and a sequence tag located between the receptor-specific portion and the first primer binding site, wherein the receptor-specific portion anneals to a different recombinant nucleic acid at a first predetermined position and extends to form a first extension product; and (ii) adding a second set of primers to the reaction mixture under primer extension conditions, wherein each primer in the second set has a receptor-specific portion, wherein the receptor-specific portion anneals to the first extension product at a second predetermined position, and wherein each primer in the second set is extended to form second extension products, wherein each second extension product comprises a first primer binding site, a sequence tag, and a recombinant nucleic acid encoding a portion of a T cell receptor chain or a B cell receptor chain; (b) amplifying the tag-nucleic acid conjugate; (c) sequencing a sample of the tag-nucleic acid conjugate to provide sequence reads, each of the sequence reads comprising a tag sequence and a recombinant nucleic acid sequence; (d) aligning sequence reads with similar tag sequences to form a group of sequence reads with the same sequence tag; (e) merging the groups of recombined nucleic acid sequences to determine clonotypes, wherein whenever at least 99.9% of the groups of recombined nucleic acid sequences are different, the sequence reads are merged into different clonotypes; and (f) detecting the presence, absence and / or presence level of a clonotype associated with the cancer of the patient in the clonotype profile, thereby detecting the minimal residual disease in the patient; wherein the recombinant nucleic acid comprises recombinant nucleic acids encoding TCRβ, TCRδ and TCRγ chains, and wherein the primers of the first group and the primers of the second group comprise primers flanking the region of the recombinant nucleic acid encoding the VDJ region of TCRβ and TCRδ and primers flanking the VJ region of TCRγ; or wherein the recombinant nucleic acid comprises recombinant nucleic acid encoding IgH and IgK, and wherein the primers of the first group and the primers of the second group comprise primers flanking the recombinant nucleic acid region encoding the VDJ region of IgH, the DJ region of IgH, and the VJ region of IgK; wherein the use further comprises the step of removing the unextended primers in the first group and the second group from the reaction mixture, wherein the removing step comprises adding an enzyme having exonuclease activity to the reaction mixture; wherein after melting the first extension product, the annealing and extension of the first group of primers are repeated; wherein after melting the second extension product, the annealing and extension of the second group of primers are repeated; wherein the sequence tag is a mosaic tag, wherein the mosaic tag comprises alternating constant and variable regions; wherein the first predetermined position is region V; wherein the receptor-specific portion of each of the primers of the first set anneals to a different non-overlapping position in the V region; wherein the second predetermined position is region J or region C; wherein the receptor-specific portion of each of the primers of the second set anneals to a different non-overlapping position in the J region; wherein the receptor-specific portion of each of the primers of the second set anneals to a single primer binding site in the J region.

2. The method of claim 1, wherein the amplification comprises performing a polymerase chain reaction in the reaction mixture to form an amplicon, the polymerase chain reaction using a forward primer specific for the first primer binding site and a reverse primer specific for the second set of primers.

3. The use according to claim 1, wherein the primers of the first group or the second group include at least one set of nested primers, and the at least one set of nested primers is specific to multiple different primer binding sites in the V region of the IgH chain. 4 . The use according to claim 2 , wherein the forward primer and / or the reverse primer comprises a 5′ non-complementary end containing a sample tag.

Images