Quantitative amplicon sequencing for multiplex copy number variation detection and allele ratio quantification

Abstract

This article provides a method for quantitative amplicon sequencing, for each chain of the targeted genomic locus in the oligonucleotide barcode sequence-labeled DNA sample by polymerase chain reaction, and amplification of the genomic region for high-throughput sequencing. By quantifying the frequency of the extra copy of each gene, these methods can be used to simultaneously detect the copy number variation (CNV) in a group of target genes. In addition, these methods use multiple PCR to quantify the allele ratios of the different genetic identities of the targeted genomic locus. In addition, these methods provide the quantification of mutation detection and variant allele frequency.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims the benefit of priority to U.S. Provisional Application No. 62 / 788,375, filed on January 4, 2019, the entire contents of which are incorporated herein by reference.

[0003] Statement Regarding Federally Sponsored Research

[0004] This invention was made with government support under Grant No. R01HG008752 awarded by the National Institutes of Health. The government has certain rights in this invention.

[0005] References to sequence listings

[0006] This application contains a sequence listing, which has been submitted in ASCII format via EFS-WEB and is hereby incorporated by reference in its entirety. The ASCII copy was created on November 26, 2019, named RICEP0058WO_ST25.TXT, and is 145.6KB in size. Technical Field

[0007] The present invention relates generally to the fields of molecular biology and medicine. More specifically, it relates to compositions and methods for multiplex copy number variation detection and allele ratio quantification using quantitative amplicon sequencing. Background Art

[0008] Copy number variations (CNVs) are important cancer biomarkers that contribute to the development and progression of cancer. They are present in a large proportion of tumors, ranging from 3% to 98% depending on the cancer type. Many CNVs confer sensitivity or resistance to targeted therapies, for example, MET amplification increases sensitivity to MET TKIs in non-small cell lung cancer, while PTEN loss confers resistance to BRAF inhibitors in melanoma. In a tumor sample, CNVs of a specific gene may be present in only a small fraction (<10%) of cells due to tumor heterogeneity and normal cell contamination.

[0009] Unlike mutations and indels, CNVs do not have unique sequences, so detecting CNVs requires accurate quantification. This quantification is difficult due to the randomness of DNA molecule sampling. For example, the standard deviation (σ) of sampling 1200 molecules per locus (i.e., 1200 haploid genome copies from 600 normal cells, 4ng genomic DNA) can be estimated using the Poisson distribution: corresponds to 3% of the number of molecules. In this case, it is impossible to detect 1% of extra copies. In theory, increasing the number of input molecules or analyzing more loci could similarly reduce the variance, and σ could be estimated as If the genome copy number or number of loci increases 100-fold, σ will decrease to 0.3%, and 1% of additional copies can be detected.

[0010] The current standard method for CNV detection in molecular diagnostics is in situ hybridization (ISH), which can determine CNV status based on the observation of a small number of cells. However, due to the limited number of distinguishable colors in fluorescence and bright field microscopy, ISH technology lacks the ability to analyze multiple genomic regions simultaneously. In addition, ISH is a complex process that needs to be performed by specialized laboratories, so it cannot be widely adopted.

[0011] Another approach to CNV detection is droplet digital PCR (ddPCR), a PCR-based method for absolute quantification of DNA molecules. However, its limit of detection (LoD) for CNVs is approximately 20% extra copies across a large number of replicates. Like ISH, ddPCR cannot be multiplexed due to the limited number of fluorescent channels. Microarray-based methods, including array comparative genomic hybridization and SNP arrays, are highly multiplexed methods for screening large CNVs and aneuploidies. However, they perform poorly in detecting smaller CNVs <40 kb or low-frequency CNVs with <30% extra copies.

[0012] Next-generation sequencing (NGS) is a high-throughput technology that has rapidly decreased in cost over the past decade. NGS is popular in the field of cancer molecular diagnostics. Highly multiplexed mutation detection with LoD < 0.1% variant allele frequency has been achieved and commercialized on NGS platforms. However, the LoD of current NGS methods for CNV detection is not that good: whole exome sequencing (WES) has been used for CNV discovery at a level of approximately 30% extra copies, but is expensive and requires more NGS reads (with proportionally increased costs) to achieve lower LoDs. Smaller hybrid capture panels, such as the FoundationOne commercial panel, can achieve LoDs of approximately 30% extra copies at a lower cost.

[0013] In NGS panels for diagnostics, target enrichment is required to reduce NGS reads wasted on irrelevant genomic regions. Two popular target enrichment methods are hybrid capture and multiplex PCR. Most current NGS-based CNV panels are based on hybrid capture, which means that the target region is captured by biotinylated nucleic acid probes and separated from the rest of the genome using streptavidin magnetic beads. The on-target rate of hybrid capture panels is low when the panel size is small, so most panels are >100kb (i.e., >1000 probes or loci); this is due to nonspecific binding of unwanted DNA on the bead surface, probes, and captured targets. Due to the large number of loci, the coverage of hybrid capture panels is not uniform: the 95% and 5% percentile loci differ by at least 30 times, which introduces another layer of bias in quantification. Hybrid capture panels also have the problem of low conversion rate (i.e., the percentage of input molecules sequenced) due to imperfect end repair and ligation, which leads to biased sampling processing and causes variation. Summary of the invention

[0014] This article provides a method for quantitative amplicon sequencing, which is used to mark each strand of a targeted genomic locus in a DNA sample with an oligonucleotide barcode sequence by polymerase chain reaction, and amplify the genomic region for high-throughput sequencing. By quantifying the frequency of extra copies of each gene, these methods can be used to simultaneously detect copy number variations (CNVs) in a group of target genes. In addition, these methods use multiplex PCR to quantify the allele ratios of different genetic identities of the targeted genomic locus.

[0015] In one embodiment, the present invention provides a method for preparing a genomic DNA targeted region for high-throughput sequencing, the method comprising: (a) obtaining a genomic DNA sample; (b) amplifying at least a portion of the genomic DNA sample by performing two PCR cycles using: (i) a first oligonucleotide comprising, from 5′ to 3′, a first region, a second region having a length of 0 to 50 nucleotides (e.g., 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 4, 5, 6, 7, 8, 9, 10, 11, or 12 degenerate nucleotides), and a fourth region comprising a sequence complementary to the first target genomic DNA region; and (ii) a second oligonucleotide having a length of 0 to 50 nucleotides from 5′ to 3′ including the fifth region. (c) comprising a sixth region (e.g., 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49 or 50 nucleotides) of the target genomic DNA region and a seventh region comprising a sequence complementary to the second target genomic DNA region; and (c) using a sequence greater than that used in step (b) The annealing temperature of step (b) is 0-10°C (e.g., 1-10, 2-10, 3-10, 4-10, 5-10, 1-9, 1-8, 1-7, 1-6, 1-5, 2-9, 2-8, 2-7°C or any range or value derivable therein) higher than the annealing temperature of step (b), and at least three PCR cycles are performed to amplify the product of step (b) using: (i) a third oligonucleotide comprising a sequence capable of hybridizing to the reverse complement sequence of at least a portion of the first region; and (ii) a fourth oligonucleotide comprising a sequence capable of hybridizing to the reverse complement sequence of at least a portion of the fifth region;and (d) amplifying the product of step (c) by performing at least one PCR cycle using a fifth oligonucleotide comprising, from 5′ to 3′, an eighth region, a ninth region between 0 and 50 nucleotides in length (e.g., 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, or 50 nucleotides), and a tenth region comprising a sequence complementary to a third target genomic DNA region, wherein the third target genomic DNA region is closer to the first target genomic DNA region by at least one nucleotide than the second target genomic DNA region.;

[0016] In some aspects, the method is a method for preparing 1 to 10,000 targeted regions (e.g., at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75, 100, 250, 500, 750, 1000, 2000, 3000, 4000, or 5000 and at most 10,000, 9000, 8000, 7000, 6000, 5000, 4000, 3000, 2000, 1000, 750, 500, 250, 100, 75, or 50 targeted regions, or any range or value derivable therein) of genomic DNA for high throughput sequencing. In some aspects, the third region is a unique molecule identifier (UMI). In some aspects, the third target genomic DNA region is closer to the first target genomic DNA region by 1-10 (e.g., 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10) bases than the second target genomic DNA region. In some aspects, the first region and the eighth region are universal primer binding sites. In some aspects, the first region and the eighth region include complete or partial NGS adapter sequences. In some aspects, the fifth region includes a sequence that cannot be found in the human genome. In some aspects, the fifth region includes a sequence different from the NGS adapter sequence. In some aspects, the melting temperature of the first region and the fifth region is 0-10°C (e.g., 1-10, 2-10, 3-10, 4-10, 5-10, 1-9, 1-8, 1-7, 1-6, 1-5, 2-9, 2-8, 2-7°C or any range or value that can be derived therefrom) higher than the melting temperature of the fourth region and the seventh region. In some aspects, the degenerate nucleotides in the third region are each independently one of A, T or C. In some aspects, none of the degenerate nucleotides in the third segment are G. In some aspects, there is a population of first oligonucleotides, each having a unique third segment.

[0017] In some aspects, the method further includes the product of purification step (c). In some aspects, purification includes SPRI purification or column purification. In some aspects, the method further includes the product of purification step (d). In some aspects, purification includes SPRI purification or column purification. In some aspects, the method further includes (e) using primers hybridized with the first region and the eighth region to amplify the product of step (d) by PCR, wherein the primer comprises an index sequence for next generation sequencing. In some aspects, the method further includes the product of purification step (e). In some aspects, purification includes SPRI purification or column purification. In some aspects, the method further includes (f) high-throughput DNA sequencing of the product of step (e). In some aspects, high-throughput DNA sequencing includes next generation sequencing.

[0018] In some aspects, the first target genomic DNA region and the second target genomic DNA region are on the opposite strand of genomic DNA. In some aspects, the first target genomic DNA region and the second target genomic DNA region are separated by 40 nucleotides to 500 nucleotides (for example, 40, 45, 50, 55, 60, 65, 70, 75, 80, 90, 100, 125, 150, 175, 200, 225, 250, 275, 300, 325, 350, 375, 400, 425, 450, 475 or 500 nucleotides, or any value that can be derived therein). In some aspects, step (b) includes an extension time of about 30 minutes (for example 27, 28, 29, 30, 31, 32 or 33 minutes). In some aspects, step (c) includes an extension time of about 30 seconds (for example 27, 28, 29, 30, 31, 32 or 33 seconds). In some aspects, step (d) comprises an extension time of about 30 minutes (eg, 27, 28, 29, 30, 31, 32, or 33 minutes).

[0019] In one embodiment, the present invention provides a method for quantifying the frequency of extra copies (FEC) of at least one target gene, the method comprising: (a) obtaining a genomic DNA sample; (b) preparing genomic DNA for high-throughput sequencing according to the method of any one of the embodiments of the present invention, wherein the sequences of the fourth region, the seventh region, and the tenth region hybridize with at least one target gene; (c) performing high-throughput sequencing according to the method of any one of the embodiments; and (d) calculating the FEC of at least one target gene based on the sequencing information obtained in step (c).

[0020] In some aspects, the method is a method of quantifying the FEC of a set of target genes, wherein the set of target genes comprises 2 to 1000 target genes (e.g., at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75, 100, 250, 500, or 750, and at most 1,000, 900, 800, 750, 700, 650, 600, 550, 500, 450, 400, 350, 300, 250, 200, 150, 100, 75, 50, 25, 20, 15, 10, 9, 8, 7, 6, 5, 4, or 3 targeted regions, or any range or value derivable therein). In some respects, the first oligonucleotide colony, the second oligonucleotide colony and the fifth oligonucleotide colony are used to carry out step (b), wherein a part of each colony in the first, second and fifth oligonucleotide colonies respectively comprises the fourth, seventh and tenth regions complementary with one in the group of target gene.In some respects, each in the fourth, seventh and tenth regions is included in the sequence that only finds once in the human genome.In some respects, each first oligonucleotide hybridized with a target gene is compared with each other first oligonucleotide hybridized with the same target gene, with unique third region.In some respects, the first oligonucleotide, the second oligonucleotide and the fifth oligonucleotide respectively comprising the fourth, seventh and tenth regions complementary with reference gene are carried out step (b). In some aspects, step (b) prepares a portion of each target gene or reference gene for high-throughput sequencing, wherein the portion is between 40 nucleotides and 500 nucleotides in length (e.g., 40, 45, 50, 55, 60, 65, 70, 75, 80, 90, 100, 125, 150, 175, 200, 225, 250, 275, 300, 325, 350, 375, 400, 425, 450, 475, or 500 nucleotides, or any value derivable therein). In some aspects, FEC is defined as:

[0021]

[0022] In some aspects, step (d) includes: (i) aligning the NGS reads to the targeted portion of each target gene and grouping the NGS reads into subsets based on the loci to which they align; (ii) partitioning the NGS reads based on their UMI sequences at each locus so that all NGS reads carrying the same UMI sequence are grouped into a UMI family; (iii) removing UMI families resulting from PCR errors or NGS errors; (iv) calculating the number of unique UMI sequences at each locus; and (v) calculating the FEC based on the number of unique UMI sequences at each locus in each target gene and reference gene. In some aspects, step (d)(iii) includes removing UMI sequences that do not conform to the UMI degenerate base design. In some aspects, step (d)(iii) comprises removing UMI families whose UMI family size is less than Fmin, where the UMI family size is the number of reads carrying the same UMI, where Fmin is between 2 and 20 (e.g., 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20). In some aspects, step (d)(iv) comprises removing UMI sequences that differ by only one or two bases from another UMI sequence with a larger family size.

[0023] In some aspects, FEC is defined as:

[0024]

[0025] in is the sum of the number of unique UMIs of all or part of the target gene loci, u is the number of loci to be considered, and u does not exceed the total number of loci in the target gene; is the sum of the number of unique UMIs of all or part of the reference loci, v is the number of loci to be considered for a reference, v does not exceed the total number of loci in the reference; w is the number of references to be considered, w does not exceed the total number of references; and k is determined by experimental calibration. In some aspects, FEC is used to identify the copy number variation (CNV) status of a target gene.

[0026] In one embodiment, the present invention provides a method for quantifying the allele ratios of different genetic identities of at least one target genomic locus, the method comprising: (a) obtaining a genomic DNA sample; (b) preparing genomic DNA for high-throughput sequencing according to the method of any one of the embodiments of the present invention, wherein the sequences of the fourth region, the seventh region, and the tenth region are hybridized with the genomic DNA near the at least one target genomic locus; (c) performing high-throughput sequencing according to the method of any one of the embodiments; and (d) calculating the allele ratios of different genetic identities of at least one target genomic locus based on the sequencing information obtained in step (c).

[0027] In some aspects, the method is a method for quantifying allele ratios of different genetic identities of a set of target genomic loci, wherein the set of target genomic loci comprises 2 to 10,000 target genomic loci (e.g., at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75, 100, 250, 500, 750, 1,000, 2,000, 3,000, 4,000, or 5,000 and at most 10,000, 9,000, 8,000, 7,000, 6,000, 5,000, 4,000, 3,000, 2,000, 1,000, 750, 500, 250, 100, 75, or 50 target genomic loci, or any range or value derivable therein). In some aspects, step (b) is carried out using the first oligonucleotide population, the second oligonucleotide population, and the fifth oligonucleotide population, wherein a portion of each population in the first, second, and fifth oligonucleotide populations comprises the fourth, seventh, and tenth regions complementary to the genomic DNA near at least one of the group of target genomic loci, respectively. In some aspects, each of the fourth, seventh, and tenth regions includes a sequence that cannot hybridize with the non-target region of genomic DNA under the conditions of step (b). In some aspects, each first oligonucleotide hybridized with the genomic DNA near a target genomic locus has a unique third region compared to each other first oligonucleotide hybridized with the genomic DNA near the same target genomic locus. In some aspects, each target genomic locus is between 40 nucleotides and 500 nucleotides in length (e.g., 40, 45, 50, 55, 60, 65, 70, 75, 80, 90, 100, 125, 150, 175, 200, 225, 250, 275, 300, 325, 350, 375, 400, 425, 450, 475, or 500 nucleotides, or any value derivable therein).

[0028] In some aspects, step (d) includes: (i) aligning the NGS reads to the targeted genomic loci and grouping the NGS reads into subgroups based on the loci to which they are aligned; (ii) partitioning the NGS reads based on their UMI sequences at each locus so that all NGS reads carrying the same UMI sequence are grouped into a UMI family; (iii) removing UMI families resulting from PCR errors or NGS errors; (iv) calling the genetic identity of each remaining UMI family; (v) calculating the number of unique UMI sequences at each locus; and (vi) calculating allele ratios. In some aspects, step (d)(iii) includes removing UMI sequences that do not conform to the UMI degenerate base design. In some aspects, step (d)(iii) comprises removing UMI families whose UMI family size is less than Fmin, wherein the UMI family size is the number of reads carrying the same UMI, wherein Fmin is between 2 and 20 (e.g., 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20). In some aspects, step (d)(iii) comprises removing UMI sequences that differ by only one or two bases from another UMI sequence having a larger family size. In some aspects, step (d)(iv) comprises calling genetic identity only when at least 70% (e.g., 70%, 75%, 80%, 85%, 90%, 95%, or 98%) of the reads of the UMI family are identical at the target genetic locus. In some aspects, the allele ratio is defined as R allele =N1 / N2, where N1 is the number of unique UMIs of the first genetic identity and N2 is the number of unique UMIs of the second genetic identity.

[0029] In some aspects, step (d)(iv) includes identifying a consensus sequence for each UMI family. In some aspects, the consensus sequence is the sequence that occurs the most times in the UMI family. In some aspects, step (d)(iv) further includes comparing the consensus sequence to the wild-type sequence of the locus to identify a mutation in the consensus sequence. In some aspects, the method further includes calculating a variant allele frequency (VAF) for the identified mutation. In some aspects, the VAF for the identified mutation is defined as the number of UMI families with the mutation / the total number of UMI families.

[0030] As used herein, "substantially free" with respect to a particular component is used herein to mean that any particular component is not intentionally formulated into the composition and / or is present only as a contaminant or in trace amounts. Thus, the total amount of the particular component resulting from any accidental contamination of the composition is much less than 0.05%, preferably less than 0.01%. Most preferred are compositions in which the particular component cannot be detected using standard analytical methods.

[0031] As used in this specification, "a" or "an" may refer to one or more. As used in one or more claims, when used with the word "comprising", "a" or "an" may refer to one or more.

[0032] Although the present disclosure supports definitions that refer only to alternatives and "and / or", the term "or" used in the claims means "and / or" unless explicitly stated to refer only to alternatives or the alternatives are mutually exclusive. As used herein, "another" can mean at least a second or more.

[0033] Throughout this application, the term "about" is used to indicate a value that includes inherent error variations of the device, the method used to determine the value, or variations that exist between subjects.

[0034] Through the following detailed description, other purposes, features and advantages of the present invention will become apparent. However, it should be understood that these detailed descriptions and specific examples, while indicating preferred embodiments of the present invention, are only for demonstration, because various changes and improvements within the spirit and scope of the present invention will be apparent to those skilled in the art based on this detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The following drawings constitute part of this specification and are included to further illustrate certain aspects of the present invention. The present invention may be better understood by reference to one or more of the drawings in conjunction with the description of the specific embodiments given herein.

[0036] Figure 1.Schematic diagram of QASeq primer design and experimental workflow.Each primer set contains 3 different oligonucleotides: a specific forward primer (SfP), a specific reverse primer A (SrPA), and a specific reverse primer B (SrPB). Each QASeq combination requires only one universal forward primer (UfP) and one universal reverse primer (UrP). In UfP or UrP, there can be additional bases at the 5′ end of region 1 or region 5. For the recommended workflow, the DNA sample is first mixed with all SfPs, SrPAs, DNA polymerase, dNTPs, and PCR buffer. Two cycles of long extension PCR are performed to add UMIs on all target loci. Next, to prevent the addition of multiple UMIs on the same original molecule while amplifying the molecules, the annealing temperature is increased by about 8°C during about 7 cycles of PCR amplification using UfP and UrP (short extension, about 30s); note that the addition of UfP and UrP to the reaction is an open-tube step on the thermal cycler. After purification using SPRI magnetic beads or columns, SrPB primers, DNA polymerase, dNTPs, and PCR buffer are mixed with PCR products for adapter replacement; after 2 cycles of long extension (about 30 minutes), NGS adapters are added only to the correct PCR products, not primer dimers or nonspecific products. After purification again using SPRI magnetic beads or columns, standard NGS index PCR is performed; the library is normalized and loaded onto the Illumina sequencer.

[0037] Figure 2 .Use (H) to simulate the UMI cross-binding energy. 20 Replace (N) 20 or(SWW)6SW as UMI sequence will reduce the average cross-binding energy, indicating that there are fewer primer-dimer interactions. Here, 500 simulations were performed for each UMI pattern; in each simulation, 2 sequences consistent with the pattern were randomly generated, and 60℃ and 0.18MK were assumed. + To calculate the cross-binding ΔG° between these sequences.

[0038] Figure 3A-B. Spacers between primers and UMIs reduce PCR bias. (Figure 3A) Workflow to evaluate the importance of spacers between primers and UMIs. Three sets of primers, i.e., no spacer (set 1); 5 nt spacer between forward primer and UMI and 5 nt spacer between reverse primer and UMI (set 2); or 12 nt spacer between forward primer and UMI and 11 nt spacer between reverse primer and UMI (set 3), were used to amplify input molecules. Indexes were added prior to NGS analysis on Illumina MiSeq. (Figure 3B) Histogram of experimental UMI family size distribution for the three sets of primers. UMI sequences that did not match the UMI design pattern were removed.

[0039] Figure 4A-B .Data analysis for UMI-based CNV absolute quantification. (Figure 4A) Data analysis workflow for CNV detection. NGS reads in the FASTQ output file are analyzed to generate CNV status as results. The FEC of the target gene will be calculated as in is the sum of the number of unique UMIs of all or part of the target gene loci, and u is the number of loci to be considered; is the sum of the number of unique UMIs of all or part of the reference loci, v is the number of loci to be considered for a reference; w is the number of references to be considered; and k is determined by experimental calibration. CNV status is determined based on FEC. (Figure 4B) Definition of UMI family size and number of unique UMIs in data analysis: UMI family size is the number of reads carrying the same UMI sequence, and the number of unique UMIs is the total number of different UMI sequences at a locus.

[0040] Figure 5 .Example of experimental UMI family size distribution. Example of UMI family size distribution for 10 ERBB2 amplicons and 10 reference amplicons in the same NGS library. We used normal cell line gDNA NA18562 (purchased from Coriell) as template input for a 20-plex QASeq experiment; the input sample contained 2500 haploid genome copies. The prepared NGS library was sequenced using 1.5 million reads by Illumina MiSeq Reagent Kit v3 (150 cycles). The fractions of accepted and rejected UMIs are shown as pie charts. Of all UMIs, approximately 20% were rejected due to PCR or sequencing errors (i.e., G bases were found in poly(H)UMIs); approximately 40% were rejected due to small family size (≤3).

[0041] Figure 6 .Example of the number of unique UMIs for different loci. Example of the number of unique UMIs for each locus, corresponding to Figure 5Data shown; white bars are ERBB2 amplicons, grey bars are reference amplicons. Input samples contained 2500 haploid genome copies. Prepared NGS libraries were sequenced using 1.5 million reads by Illumina MiSeq Reagent Kit v3 (150 cycles).

[0042] Figure 7 Experimental calibration results and simulated theoretical standard deviation limits of gDNA from normal cell line NA18562. CNV ratio (σ CNV比率 ) is plotted against the number of input molecules. The LoD can be approximated as 3σ CNV比率 We performed five replicates for each input size (75, 250, 750, and 2500 haploid genome copies); the results are plotted as crosses. Simulations were performed assuming a Poisson distribution of the number of sampled molecules; due to the randomness of the sampling, the simulated σ CNV比率 (drawn as a dashed line) is the theoretical lower limit.

[0043] Figure 8A-C .Example of experimental results for CNV detection on FFPE samples.We tested 2 lung cancer FFPE slides from the same tumor, in which ERBB2 CNVs were unlikely to occur. The input extracted DNA samples contained 2500 haploid genomic copies for each NGS library. The prepared NGS libraries were sequenced using 1.5 million reads by Illumina MiSeq Reagent Kit v3 (150 cycles). (Figure 8A) Example of UMI family size distribution plotted for amplicon ERBB2_1 and reference_1; the fraction of accepted and discarded UMIs are shown as pie charts. (Figure 8B) Example of the number of unique UMIs per amplicon region. White bars are ERBB2 amplicons; gray bars are reference amplicons. (Figure 8C) CNV ratios for 2 FFPE slides from the same lung cancer tumor are plotted. Based on previous calibration data, no CNVs for ERBB2 were detected in these FFPE slides using QASeq. Mean and LoD = 3σ CNV比率 is calculated based on data from a gDNA library of a cell line with 750 genomic copies of input (see Figure 7 ), which have similar numbers of unique UMIs as FFPE samples.

[0044] Figure 9A-E.Primer dimers were reduced using the main experimental workflow. (Figure 9A) The simplest workflow we tested was a one-pot reaction: after adding the UMIs, the index primers were added directly to the reaction as an open-tube step on the thermocycler, followed by index PCR (i.e., universal PCR). The on-target rate for this workflow was low (0.5%); off-target NGS reads were primarily primer dimers. (Figure 9B) An SPRI cleanup step was added after 6 cycles of universal PCR to reduce primer dimers; the on-target rate increased to 20%. (Figure 9C) A size selection step using agarose gel was added after index PCR to further reduce primer dimers; the on-target rate increased compared to Figure 9B but remained below 50%. (Figure 9D) The main experimental workflow included adapter replacement and cleanup after universal PCR, which resulted in an average on-target rate of up to 66%. (Figure 9E) Sources of primer dimers in workflows 9A-D.

[0045] Figure 10A-C .Workflow example for NGS-free indexed PCR. (Figure 10A) Index and P5 sequences are added to 5′ of UfP; additional index and P7 sequences are added to 5′ of SrPB. The amplicon obtained from the adapter exchange contains P5, P7, and the dual index and is therefore ready for sequencing. (Figure 10B) Index and P7 sequences are added to 5′ of SrPB, and index primers are added together with SrPB in the adapter exchange step. The amplicon is ready for sequencing. (Figure 10C) Index and P5 sequences are added to 5′ of SfP; primers with P5 sequences are used as UfP in the universal PCR step. Additional index and P7 sequences are added to 5′ of SrPB. The amplicon is ready for sequencing.

[0046] Fig.11 .Variations of QASeq Primer Design and Workflow. Each primer set contains 3 different oligonucleotides: Specific forward primer (SfP), specific reverse primer A (SrPA), and specific reverse primer B (SrPB). Compared to the original design, SrPA only requires the template binding region and does not require a universal reverse primer (UrP). Each QASeq combination requires only one universal forward primer (UfP); additional bases can be added to the 5′ end of region 1 in the UfP. Compared to the original experimental workflow, the universal PCR step requires more PCR cycles; ≥10 cycles are recommended.

[0047] Figure 12A-B .Data analysis of allele ratio quantification based on QASeq. (Figure 12A) Data analysis workflow for allele ratio quantification. NGS reads in FASTQ output files were analyzed to generate allele ratios between different genetic identities. The allele ratio for each targeted locus was calculated as R 等位基因= N1 / N2, where N1 is the number of unique UMIs of the first genetic identity and N2 is the number of unique UMIs of the second genetic identity. (FIG. 12B) Genetic identity calls for each UMI family based on majority vote.

[0048] Fig.13 .Example of experimental results of CNV detection spiked into clinical FFPE samples. Two previously characterized FFPE DNA samples (1 "normal" sample and 1 "ERBB2 amplification abnormal" sample) were mixed to generate 2.5%, 5%, and 10% ERBB2 FEC samples. The ERBB2 FEC of the "normal" sample was 0%, and the ERBB2 FEC of the "ERBB2 amplification abnormal" sample was 78%. The experimental normalized FEC values ​​were plotted against the expected ERBB2 FEC. The "normal" sample was tested 5 times, and the LoD of the 100-plex CNV combination was estimated to be 3 standard deviations of the "normal" sample. CNVs in the 2.5%, 5%, and 10% ERBB2 FEC samples were successfully detected because their calculated FECs were outside the 3 standard deviation range.

[0049] Fig.14 .Bioinformatics workflow for mutation quantification using QASeq. Shown is an overview of the data processing workflow for mutation quantification.

[0050] Fig.15 . Observed number of molecules for 179-plex comprehensive panel. Input was 8.3 ng (5000 expected number of molecules) of 100% Multiplex I Wild Type cfDNA Reference Standard (Horizon Discovery). Conversion averaged 62%; 97% of the panel had >10% conversion.

[0051] Fig.16 .179-plex comprehensive combination error rate. The input was 8.3ng of 100% Multiplex I Wild Type cfDNA Reference Standard (Horizon Discovery); the same sample was tested in triplicate. The error rates of 3840 different loci (after error correction using UMI) were plotted. The highest error rates of the 3 replicates were 0.23%, 0.20%, and 0.23%, and the average error rates were 0.006%, 0.005%, and 0.005%.

[0052] Fig.17.179-plex comprehensive panel mutation quantification results. The sample used was 0.3% cfDNA reference standard (created by mixing 0.1% Multiplex I cfDNA Reference Standard and 1% Multiplex I cfDNA Reference Standard from Horizon Discovery), tested in triplicate. The experimental VAFs for the 6 mutations were generally consistent with the expected VAFs; the differences were mainly due to the randomness of sampling a small number (≤9) of mutant molecules. DETAILED DESCRIPTION

[0053] Provided herein is a method for quantitative amplicon sequencing, for labeling each chain of a targeted genomic locus in an original DNA sample with an oligonucleotide barcode sequence by polymerase chain reaction, and amplifying a genomic region for high-throughput sequencing. Also provided herein is a method for simultaneously detecting a copy number variation (CNV) in a set of target genes by quantifying the frequency of an extra copy of each gene. The disclosed method also provides quantification of allele ratios of different genetic identities of a targeted genomic locus using multiplex PCR. These methods can be used to detect CNVs of target genes in tumor samples, guide the selection of targeted therapies, and help understand the formation and progression of cancer.

[0054] The current standard approach for prenatal diagnosis of monogenic diseases is to sequence fetal genetic material obtained from invasive and risky chorionic villus sampling or amniocentesis. Genetic noninvasive prenatal testing (NIPT) for monogenic diseases is based on the circulation of fetal-derived cell-free DNA (cfDNA) in maternal plasma. Due to the presence of background maternal DNA, it becomes challenging to reliably detect changes in allele ratios caused by fetal cfDNA, especially when the maternal DNA is heterozygous at the target locus. Droplet digital PCR (ddPCR) has been used to quantify the allele ratio between mutant alleles carrying pathogenic mutations and wild-type alleles for NIPT (Lun et al., 2008), but practical feasibility is limited by technical precision and reliability. QASeq achieves absolute quantification of DNA molecules by adding unique molecular identifiers to each chain of the original input molecule and can be applied to allele ratio quantification for NIPT. Therefore, QASeq can also be used for allele ratio quantification. Allele ratio quantification aims to quantify the ratio of DNA molecules with different genetic identities. Accurate allele ratio quantification is key to NIPT for monogenic diseases such as β-thalassemia and cystic fibrosis.

[0055] I. Frequency of CNV extra copies

[0056] The frequency of extra copies of a CNV in a genomic DNA sample (FEC) is defined herein as:

[0057]

[0058] A positive value of FEC indicates amplification of the target genomic region in the sample, and a negative value of FEC indicates deletion of the target genomic region in the sample.

[0059] While QASeq can be used to quantify FEC, it does not provide information about the percentage of cells in a tumor tissue sample that contain a CNV. For example, if 1% of cells in a tumor sample contain 4 copies of ERBB2 and the remaining 99% of cells contain 2 copies, the FEC is 1%; if 0.5% of cells in a sample contain 6 copies of ERBB2 and the remaining 99.5% of cells contain 2 copies, the FEC is still 1%. In addition, QASeq does not provide information about the genomic location of the extra copies.

[0060] II. Multiplex PCR panel design

[0061] In the QASeq multiplex PCR panel, one target gene requires M (M = 1 to 1000) sets of primers, each set of primers amplifies a small non-overlapping region (40 nt to 500 nt, usually ≤ 200 nt) in the target gene region. If the panel has multiple target genes, the number of primer sets used for each gene is similar (≈M). The panel also contains a similar number (≈M) of primer sets for amplifying reference genomic regions. The reference locus is used as an internal standard for the amount of genomic DNA (gDNA) loaded, so accurate quantification of the DNA concentration in the sample is not required. At least one reference primer set can be used for each panel. Because increasing the number of input molecules or loci in the target gene can reduce random sampling variations, a larger number of primer sets can be used for each gene to improve the LoD of sample types containing less DNA; in this case, the number of reference primer sets needs to be increased proportionally.

[0062] Each primer set contains three different oligonucleotides: specific forward primer (SfP), specific reverse primer A (SrPA) and specific reverse primer B (SrPB) (see Figure 1). SfP includes regions 1, 2, 3, and 4 from 5′ to 3′. Region 4 is the template binding region; region 3 is the UMI; region 1 is a complete or partial NGS adapter; region 2 is an optionally added spacer (usually 0 to 15 nt) for uniform amplification of UMI. SrPA includes regions 5, 6, and 7 from 5′ to 3′. Region 7 is the template binding region; region 5 is a custom adapter for universal amplification (i.e., a sequence that is different from the NGS adapter and cannot be found in the human genome); region 6 is an optionally added spacer (usually 0 to 15 nt) for uniform amplification of different loci. SrPB includes regions 8, 9, and 10 from 5′ to 3′. Region 10 is a template binding region, whose 3' end is at least 1 base closer to region 4 than region 7; region 8 is a complete or partial NGS adapter; region 9 is an optionally added spacer (usually 0 to 15 nt) for uniform amplification of different loci. Each QASeq combination requires only one universal forward primer (UfP) and one universal reverse primer (UrP). UfP contains region 1, and UrP contains region 5; there can be additional bases at the 5′ end of region 1 or region 5 in UfP or UrP. The melting temperature (Tm) of template binding regions 4, 7, and 10 is approximately the same as the PCR annealing temperature, and the Tm of UfP and UrP is not lower than regions 4, 7, and 10 under experimental PCR conditions.

[0063] When designing primers, single nucleotide polymorphisms (SNPs) with significant minor allele frequencies (MAF) should be avoided in the primer binding region so that the binding affinity of the primers will not be affected by nucleotide sequence variants in different patient samples. In addition, the whole human genome nucleotide sequence should be searched to ensure that the primers are not prone to non-specific amplification of non-target regions.

[0064] In the combined example of CNVs of ERBB2 in formalin-fixed paraffin-embedded (FFPE) specimens targeting tumor samples, 10 sets of primers were designed in the ERBB2 gene region, each of which amplified amplicons of 60 to 70 nt. In addition, 10 sets of reference primers were designed, each of which amplified regions in different housekeeping genes from different chromosomes (Table 1). Primers were automatically designed using Matlab code to meet the above design principles while minimizing primer interactions. In addition, non-pathogenic SNPs with MAF>0.2% in the population were avoided. The online tool Primer-BLAST was used to ensure that each primer set had only one amplicon in the human genome. The primer sequences are shown in Table 2.

[0065] Table 1. Location of amplicons

[0066] Amplicon name chromosome Gene ERBB21~10 Chr.17 ERBB2 Reference 1 Chr.1 PSMB2 Reference 2 Chr.3 RPL32 Reference 3 Chr.5 RACK1 Reference 4 Chr.6 TBP Reference 5 Chr.9 VCP Reference 6 Chr.11 HMBS Reference 7 Chr.12 NACA Reference 8 Chr.15 B2M Reference 9 Chr.19 GPI Reference 10 Chr.20 TOP1

[0067] Table 2. Primer sequences in exemplary QASeq combinations

[0068]

[0069]

[0070]

[0071]

[0072] Table 3. Primer sequences in 179-recombination

[0073]

[0074]

[0075]

[0076]

[0077]

[0078]

[0079]

[0080]

[0081]

[0082]

[0083]

[0084]

[0085]

[0086]

[0087]

[0088]

[0089]

[0090]

[0091]

[0092]

[0093]

[0094]

[0095]

[0096]

[0097]

[0098]

[0099]

[0100]

[0101]

[0102]

[0103]

[0104]

[0105] III.UMI Design

[0106] During NGS library preparation, the PCR amplification step significantly increases quantitative variation, making it difficult to distinguish small changes in the number of original molecules. UMI technology can be used to reduce PCR bias and achieve absolute quantification of original DNA molecules. The concept of UMI is to give each original DNA molecule a different DNA sequence as a "barcode" so that the source of each NGS read can be traced based on the barcode sequence. Given enough NGS reads, the number of unique UMIs found in the NGS output can reflect the number of original DNA molecules. Previously, UMI technology was mainly used for error correction in NGS-based low-frequency mutation detection; it has also been applied to quantification. Each original molecule is uniquely labeled by using a large number of different UMI sequences; for example, using 109 different UMI sequences for 100,000 original molecules will produce <0.006% of molecules carrying duplicate UMIs.

[0107] DNA sequences containing degenerate bases, such as poly(N) (i.e., a mixture of A, T, C, or G at each position), are often used as UMI sequences. In QASeq, poly(H) (A, T, or C) is used as the UMI because it has weaker cross-binding energy compared to poly(N) or compared to a mixture of S (C or G) and W (A or T) bases, as shown in simulations ( Figure 2 ). (H)20 Contains 3.5×10 9 different sequences, which is sufficient for 100,000 molecules as input; (H) 15 Contains 1.4×10 7 different sequences, which is sufficient for 6,000 molecules as input.

[0108] IV. Spacer to reduce PCR bias

[0109] PCR efficiency varies between amplicons of different sequences. Since UMIs are composed of many different sequences, spacers between primers and variable UMI regions can be used to achieve more uniform PCR efficiency.

[0110] NGS was performed to evaluate the effect of the spacer on PCR bias (Figure 3A). The template molecule has two adapters for amplification at the 5' and 3' ends and a spacer in the middle consisting of (D) 15 Three sets of primers, i.e., without any spacer (set 1); with a 5nt spacer between the forward primer and the UMI and a 5nt spacer between the reverse primer and the UMI (set 2); or with a 12nt spacer between the forward primer and the UMI and an 11nt spacer between the reverse primer and the UMI (set 3), were used to amplify the templates. Indexes were added by PCR prior to NGS analysis. (D) 15 Contains 1.4×10 7 different sequences. Since the number of input template molecules is much lower than the number of possible sequences, each unique UMI sequence has only 1 copy before amplification. All NGS reads carrying the same UMI are assumed to come from the same molecule. Therefore, the UMI family size (i.e., the number of reads carrying the same UMI) is an indicator of PCR efficiency.

[0111] The UMI family size distribution was compared to assess the significance of the spacer on PCR bias (Figure 3B). A more uniform distribution was observed when the spacer between the primer and the UMI was longer. In primer set 3, where the spacer lengths at both ends were longer than 10 nt, a significantly improved distribution was achieved.

[0112] V.QASeq Workflow

[0113] The schematic diagram of QASeq NGS library preparation workflow is shown in Figure 1.First, the DNA sample is mixed with all SfP, SrPA, DNA polymerase, dNTP and PCR buffer. Two cycles of long extension (about 30min) PCR are performed to add UMIs on all target loci. After that, each strand in a DNA molecule will carry a different UMI. Next, to prevent multiple UMIs from being added to the same original molecule when amplifying the molecule, the annealing temperature is increased by about 8°C, and amplification is performed for at least two cycles (e.g., about seven cycles) with short extension (about 30 seconds) using UfP and UrP. Adding UfP and UrP to the reaction is an open tube step on a thermal cycler. After purification using SPRI magnetic beads or columns, SrPB primers, DNA polymerase, dNTP and PCR buffer are mixed with the PCR product for adapter exchange; after at least one cycle (e.g., two cycles) of long extension (about 30min), the NGS adapter is added only to the correct PCR product, not to primer dimers or non-specific products. After further purification using SPRI magnetic beads or columns, standard NGS index PCR is performed; the libraries are normalized and loaded onto an Illumina sequencer.

[0114] All types of DNA polymerases and PCR supermixes can be used. Standard annealing, extension, and denaturation temperatures for the specific polymerase used should be followed (except for the universal PCR steps, where the annealing temperature is increased).

[0115] VI. Alternative QASeq Workflow

[0116] A workflow can be performed using SfP and SrPB, adding UMIs with two cycles of PCR followed by direct addition of index primers for index PCR. To test this, twenty sets of SfP and SrPB were used in the same reaction. The experimental on-target rate for this method was very low (0.5%), so this method may not be suitable for diagnostic NGS assays (Figure 9A). Off-target NGS reads were mainly primer dimers. In a second alternative workflow, six cycles of universal PCR were performed using UfP and Urp, followed by a purification step. These additional steps increased the on-target rate of different libraries to 12-28% (average on-target rate = 20%) (Figure 9B). A third alternative workflow based on the second alternative workflow was tested. To this end, a size selection step using agarose gel was added after the index PCR to further reduce primer dimers. The experimental average on-target rate increased to 42%, but was still below 50% (Figure 9C). Primer dimer reduction was achieved using the main experimental workflow, which included adapter replacement and purification after universal PCR, and resulted in a high average on-target rate of 66% (Figure 9D). One source of primer dimers in the above workflow is shown in Figure 9E. If the 3′ portion of SfP binds to SfPB, or the 3′ portion of SfPB binds to SfP, dimer chains with universal regions at the 5′ and 3′ ends can be generated, allowing amplification in the universal or indexed PCR steps.

[0117] The primary workflow includes a final index PCR step to add the index sequence and the sequencer's P5 / P7 sequences to the ends of the amplicons; however, there are alternative workflows that add these sequences during the UMI addition, universal PCR, or adapter replacement steps, thus eliminating the need for an index PCR step. Figure 10A-C Three examples are shown. First, the index and P5 sequence are added to the 5′ of UfP; the additional index and P7 sequence are added to the 5′ of SrPB. The amplicon obtained from the adapter replacement contains P5, P7 and dual indexes and is therefore ready for sequencing (Figure 10A). Second, the index and P7 sequence are added to the 5′ of SrPB, and this modified SrPB is mixed with the normal P5 index primer in the adapter replacement step (Figure 10B). Third, the index and P5 sequence are added to the 5′ of SfP; the primer with the P5 sequence is used as UfP in the universal PCR step. The additional index and P7 sequence are added to the 5′ of SrPB (Figure 10C).

[0118] An alternative QASeq primer design and workflow is shown in Fig.11.Each primer set contains three different oligonucleotides: a specific forward primer (SfP), a specific reverse primer A (SrPA), and a specific reverse primer B (SrPB). SfP includes regions 1, 2, 3, and 4 from 5′ to 3′. Region 4 is the template binding region; region 3 is the UMI; region 1 is a complete or partial NGS adapter; region 2 is an optionally added spacer (0-15 nt) for uniform amplification of UMIs. SrPA contains region 5, which is the template binding region. SrPB includes regions 6, 7, and 8 from 5′ to 3′. Region 8 is the template binding region, and its 3' end is at least 1 base closer to region 4 than region 5; region 6 is a complete or partial NGS adapter; region 7 is an optionally added spacer (0-15 nt) for uniform amplification of different loci. Each QASeq combination requires only one universal forward primer (UfP), which includes region 1; there can be additional bases at the 5' end of region 1 in UfP. The melting temperature (Tm) of template binding regions 4, 5, and 8 is approximately the same as the PCR annealing temperature, and the Tm of UfP under the experimental PCR conditions is not lower than regions 4, 5, and 8. Compared with the original design, SrPA only requires the template binding region and does not require a universal reverse primer (UrP). In the experimental workflow, more PCR cycles (e.g., at least 10 cycles) are required in the universal PCR step under this alternative primer design.

[0119] VII. Data Analysis Workflow

[0120] A schematic diagram of the data analysis workflow for CNV detection is shown in Figure 4A. First, the raw NGS reads are aligned to the amplicon regions; optional adapter trimming can be performed before alignment. Unaligned reads are discarded, and aligned reads are grouped according to the loci to which they are aligned.

[0121] Then, all reads aligned to the same locus are further divided by the UMI sequence, that is, reads carrying the same UMI are grouped into a UMI family. The UMI family size is the number of reads carrying the same UMI, and the number of unique UMIs is the total number of different UMI sequences at a locus (Figure 4B). Next, all unique UMI families that may be the result of PCR or NGS errors are removed. For example, UMI sequences that are inconsistent with the designed UMI pattern (e.g., the G base found in the poly(H) UMI sequence) are errors and should be removed. In addition, if two UMI sequences differ by only 1-2 bases, one family with a smaller UMI family size may have mutated from the other family and can therefore be optionally removed. After removing UMI errors, family sizes < F are also removed. min UMI family. min It is determined based on the distribution of UMI family sizes. In most cases, F min= 4. The number of unique UMIs (N) after removing the UMIs is used in the next step.

[0122] The FEC of the target gene can be calculated as:

[0123]

[0124] in is the sum of the number of unique UMIs of all or part of the target gene loci, u is the number of loci to be considered, and u does not exceed the total number of loci in the target gene; is the sum of the number of unique UMIs of all or part of the reference loci, v is the number of loci to be considered for a reference, v does not exceed the total number of loci in the reference; w is the number of references to be considered, w does not exceed the total number of references; and k is determined by experimental calibration. Before testing the QASeq combination on clinical samples, calibration experiments were performed on DNA samples with well-characterized CNV status of target genes. gDNA extracted from normal and cancer cell lines with CNV status characterized by ddPCR can be used for calibration. The FEC of normal calibration samples should be 0. The LoD of the assay is also determined by the calibration experiment; LoD is the minimum frequency of extra copies that can be detected by the assay. When testing clinical samples, the FEC of the target gene will be used to infer the CNV status; if FEC>LoD, it is inferred that the sample contains an amplification of the target gene; if FEC≤LoD, it is inferred that the sample contains a deletion of the target gene.

[0125] VIII. Quantification of allele ratios

[0126] QASeq can be used to quantify allele ratios of different genetic identities at 1 to 10,000 genomic loci using multiplex PCR. The experimental workflow of designing multiplex PCR panels for targeted genomic loci, labeling each strand of the targeted genomic loci with an oligonucleotide barcode sequence by PCR, and then amplifying the genomic region for high-throughput sequencing is similar to that of CNV detection.

[0127] A schematic diagram of the data analysis workflow for allele ratio quantification is shown in FIG12A . First, the raw NGS reads are aligned to the amplicon regions; optional adapter trimming can be performed before alignment. Unaligned reads are discarded, and aligned reads are grouped by the loci to which they are aligned. At each locus, the NGS reads are divided by the UMI sequence; all NGS reads carrying the same UMI sequence are grouped into a UMI family. As described in the Data Analysis Workflow section, unique UMI families with errors in the UMI are removed, which may be the result of PCR or NGS errors.

[0128] The genetic identity (wild type or mutant) of each remaining UMI family is called based on a majority vote; genetic identity requires support from at least 70% of the members (reads) in the same UMI family. Taking Figure 12B as an example, for a UMI family with a UMI family size = 7, all 7 reads share the same UMI sequence (shown as a 2D barcode). The genetic identity at the target locus, 6 reads are "A" and 1 read is "G". Since more than 70% of the reads in the UMI family support "A", the genetic identity of the UMI family is called "A". The 1 read corresponding to "G" is the result of a PCR or NGS error. UMI families that do not have more than 70% of the reads supporting a common genetic identity are discarded.

[0129] Next, the number of unique UMIs N (the total number of different UMI sequences at a locus) is calculated for each different genetic identity at the targeted locus; N represents the number of original chains. The allele ratio of the target locus is calculated as R 等位基因 =N1 / N2, where N1 is the number of unique UMIs of the first genetic identity and N2 is the number of unique UMIs of the second genetic identity.

[0130] IX. Definitions

[0131] As used herein, "amplification" refers to any in vitro method for increasing the number of copies of one or more nucleotide sequences. Nucleic acid amplification results in the incorporation of nucleotides into DNA or RNA. As used herein, an amplification reaction can consist of multiple rounds of DNA replication. For example, a PCR reaction can contain 30 to 100 "cycles" of denaturation and replication.

[0132] "Polymerase chain reaction" or "PCR" refers to a reaction in vitro for amplifying a specific DNA sequence by simultaneous primer extension of complementary strands of DNA. In other words, PCR is a reaction for making multiple copies or replicates of a target nucleic acid flanking primer binding sites, the reaction comprising one or more repetitions of the following steps: (i) denaturing the target nucleic acid, (ii) annealing primers to the primer binding sites, and (iii) extending the primers by a nucleic acid polymerase in the presence of nucleoside triphosphates. Typically, the reaction is cycled through different temperatures optimized for each step in a thermal cycler. The specific temperatures, duration of each step, and rate of change between steps depend on a number of factors well known to those of ordinary skill in the art, such as those described in McPherson et al., eds., PCR: A Practical Approach and PCR2: A Practical Approach (IRL Press, Oxford, 1991 and 1995, respectively).

[0133] "Primer" refers to a natural or synthetic oligonucleotide that, when 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. The nucleotide sequence added during the extension process depends on the sequence of the template polynucleotide. Typically, the primer is extended by a DNA polymerase. The length of the primer is typically compatible with its use in the synthesis of primer extension products, and the length is typically in the range of 8 to 100 nucleotides, for example, the length is between 10 to 75, 15 to 60, 15 to 40, 18 to 30, 20 to 40, 21 to 50, 22 to 45, 25 to 40, etc., more typically in the range of 18 to 40, 20 to 35, 21 to 30 nucleotides, and any length between the ranges. The length of a typical primer can range from 10 to 50 nucleotides, such as 15 to 45, 18 to 40, 20 to 30, 21 to 25, etc., and any length between the ranges. In some embodiments, the length of the primer is generally no greater than about 10, 12, 15, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 35, 40, 45, 50, 55, 60, 65, or 70 nucleotides.

[0134] As used herein, "incorporated" means becoming a part of a nucleic acid polymer.

[0135] As used herein, the term "in the absence of exogenous manipulation" refers to modifying a nucleic acid molecule without changing the solution in which the nucleic acid molecule is modified. In specific embodiments, it occurs without manual manipulation or a machine that changes the solution conditions, which may also be referred to as buffer conditions. However, temperature changes may occur during the modification process.

[0136] A "nucleoside" is a combination of a base sugar, i.e., a nucleotide lacking a phosphate. It is well recognized in the art that there is a certain interchangeability in the use of the terms nucleoside and nucleotide. For example, the nucleotide deoxyuridine triphosphate, i.e., dUTP, is a deoxyribonucleoside triphosphate. After incorporation into DNA, it serves as a DNA monomer, formally as deoxyuridine monophosphate, i.e., dUMP or deoxyuridine monophosphate. It can be said that dUTP can be incorporated into DNA even if there is no dUTP portion in the resulting DNA. Similarly, it can be said that deoxyuridine can be incorporated into DNA even if it is only a portion of a substrate molecule.

[0137] As used herein, "nucleotide" is a term of art and refers to a combination of an alkali sugar phosphate. Nucleotide is a nucleic acid polymer, i.e., the monomeric unit of DNA and RNA. The term includes ribonucleotide triphosphates, such as rATP, rCTP, rGTP or rUTP, and deoxyribonucleotide triphosphates, such as dATP, dCTP, dUTP, dGTP or dTTP.

[0138] The term "nucleic acid" or "polynucleotide" generally refers to at least one molecule or chain of DNA, RNA, DNA-RNA chimera, or a derivative or analog thereof, the at least one molecule or chain comprising at least one nucleobase, such as a naturally occurring purine or pyrimidine base found in DNA (e.g., adenine "A", guanine "G", thymine "T", and cytosine "C") or a naturally occurring purine or pyrimidine base found in RNA (e.g., A, G, uracil "U", and C). The term "nucleic acid" encompasses the terms "oligonucleotide" and "polynucleotide". As used herein, "oligonucleotide" is collectively and interchangeably referred to as two terms of the art, "oligonucleotide" and "polynucleotide". Please note that although oligonucleotide and polynucleotide are different technical terms, there is no exact dividing line between them, and they are used interchangeably herein. The term "adapter" can also be used interchangeably with the terms "oligonucleotide" and "polynucleotide". In addition, the term "adapter" can represent a linear adaptor (single-stranded or double-stranded) or a stem-loop adaptor. These definitions generally refer to at least one single-stranded molecule, but in certain embodiments, at least one additional strand that is partially, substantially, or completely complementary to the at least one single-stranded molecule will also be encompassed. Thus, a nucleic acid may encompass at least one double-stranded molecule or at least one triple-stranded molecule that contains one or more complementary strands or "complements" to a particular sequence of strands comprising the molecule. As used herein, a single-stranded nucleic acid may be indicated by the prefix "ss", a double-stranded nucleic acid by the prefix "ds", and a triple-stranded nucleic acid by the prefix "ts".

[0139] "Nucleic acid molecule" or "nucleic acid target molecule" refers to any single-stranded or double-stranded nucleic acid molecule, including standard classical bases, super-modified bases, non-natural bases or any combination of their bases. For example, but not limited to, the nucleic acid molecule contains the four classical DNA bases-adenine, cytosine, guanine and thymine, and / or the four classical RNA bases-adenine, cytosine, guanine and uracil. When the nucleoside contains a 2'-deoxyribose group, uracil can replace thymine. Nucleic acid molecules can be converted from RNA to DNA, or from DNA to RNA. For example, but not limited to, reverse transcriptase can be used to generate mRNA into complementary DNA (cDNA), and RNA polymerase can be used to generate DNA into RNA. Nucleic acid molecules can be of biological or synthetic origin. Examples of nucleic acid molecules include genomic DNA, cDNA, RNA, DNA / RNA hybrids, amplified DNA, pre-existing nucleic acid libraries, etc. Nucleic acids can be obtained from human samples, such as blood, serum, plasma, cerebrospinal fluid, cheek scrapings, biopsies, semen, urine, feces, saliva, sweat, etc. Nucleic acid molecules can be subjected to various treatments, such as repair treatments and fragmentation treatments. Fragmentation treatments include mechanical, sonic and hydrodynamic shearing. Repair treatments include gap repair by extension and / or ligation, polishing to produce flat ends, removal of damaged bases, such as deaminated, derivatized, abasic or cross-linked nucleotides, etc. Target nucleic acid molecules can also be chemically modified (e.g., bisulfite conversion, methylation / demethylation), extended, amplified (e.g., PCR, isothermal, etc.), etc.

[0140] "Complementary" nucleic acids or "complementary sequences" are nucleic acids that are capable of base pairing according to standard Watson-Crick, Hoogsteen or reverse Hoogsteen binding complementarity rules. As used herein, the terms "complementary" or "complementary sequences" may refer to substantially complementary nucleic acids, as may be assessed by the same nucleotide comparison described above. The term "substantially complementary" may refer to a nucleic acid comprising at least one sequence of contiguous nucleobases or semi-contiguous nucleobases (if one or more nucleobase portions are not present in the molecule) that is capable of hybridizing to at least one nucleic acid strand or duplex, even if less than all nucleobases that do not base pair with corresponding nucleobases. In certain embodiments, a "substantially complementary" nucleic acid contains at least one sequence wherein about 70%, about 71%, about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, about 100%, and any range of nucleobase sequences therein are capable of base pairing with at least one single-stranded or double-stranded nucleic acid molecule during hybridization. In certain embodiments, the term "substantially complementary" refers to at least one nucleic acid that can hybridize to at least one nucleic acid strand or duplex under stringent conditions. In certain embodiments, a "partially complementary" nucleic acid comprises at least one sequence that can hybridize to at least one single-stranded or double-stranded nucleic acid under low stringency conditions, or comprises at least one sequence in which less than about 70% of the nucleobase sequence is capable of base pairing with at least one single-stranded or double-stranded nucleic acid molecule during hybridization.

[0141] The term "non-complementary" refers to nucleic acid sequences that lack the ability to form at least one Watson-Crick base pair through specific hydrogen bonding.

[0142] As used herein, the term "degenerate" refers to a nucleotide or a series of nucleotides, wherein the identity can be selected from a variety of nucleotide selections, rather than a definite sequence. In a specific embodiment, two or more different nucleotides can be selected. In a further specific embodiment, the selection of nucleotides at a specific position includes being selected from only purines, only pyrimidines, or being selected from non-pairing purines and pyrimidines.

[0143] "Sample" refers to material obtained or isolated from a fresh or preserved biological sample or synthetically produced source, which contains a target nucleic acid. A sample may include at least one cell, fetal cell, cell culture, tissue specimen, blood, serum, plasma, saliva, urine, tears, vaginal secretions, sweat, lymph, cerebrospinal fluid, mucosal secretions, peritoneal fluid, ascites, fecal material, body exudates, umbilical cord blood, chorionic villi, amniotic fluid, embryonic tissue, multicellular embryos, lysates, extracts, solutions, or reaction mixtures suspected of containing target immune nucleic acids. Samples may also include non-human sources, such as non-human primates, rodents and other mammals, other animals, plants, fungi, bacteria, and viruses.

[0144] As used herein with respect to nucleotide sequences, "substantially known" refers to having enough sequence information to allow the preparation of nucleic acid molecules, including amplification thereof. Although some portions of the adapter sequence are random or degenerate in some embodiments, this is generally about 100%. Therefore, in particular embodiments, substantially known refers to about 50% to about 100%, about 60% to about 100%, about 70% to about 100%, about 80% to about 100%, about 90% to about 100%, about 95% to about 100%, about 97% to about 100%, about 98% to about 100%, or about 99% to about 100%.

[0145] X. Further processing of target nucleic acid

[0146] A. DNA Amplification

[0147] Many template-dependent processes can be used to amplify nucleic acids present in a given template sample. One of the best-known amplification methods is the polymerase chain reaction (PCR). TM ), which is described in detail in U.S. Pat. Nos. 4,683,195, 4,683,202 and 4,800,159 and Innis et al., 1990, each of which is incorporated herein by reference in its entirety. In short, two synthetic oligonucleotide primers complementary to two regions of the template DNA to be amplified (one for each chain) are added to the template DNA (which does not need to be pure) in the presence of excess deoxynucleotides (dNTPs) and a thermostable polymerase, such as Taq (Thermus aquaticus) DNA polymerase. In a series of (usually 30-35) temperature cycles, the target DNA is repeatedly denatured (about 90°C), annealed with the primers (usually at 50-60°C), and daughter chains are extended from the primers (72°C). When daughter chains are created, they serve as templates in subsequent cycles. Therefore, the template region between the two primers is amplified exponentially, rather than linearly.

[0148] B. DNA Sequencing

[0149] Methods for sequencing adapter-ligated fragment libraries are also provided. Any technique known to those skilled in the art for sequencing nucleic acids can be used in the methods of the present disclosure. DNA sequencing techniques include the classic dideoxy sequencing reaction (Sanger method), which uses labeled terminators or primers and gel separation in a plate or capillary; sequencing by synthesis using reversibly terminated labeled nucleotides; pyrophosphate sequencing; 454 sequencing; allele-specific hybridization with labeled oligonucleotide library probes; sequencing by synthesis using allele-specific hybridization with labeled clone libraries, followed by ligation; real-time monitoring of the incorporation of labeled nucleotides during the polymerization step; and SOLiD sequencing.

[0150] Methods compatible with Illumina sequencing can be used (e.g. Nextera TM DNA sample preparation kit) to generate nucleic acid libraries, as well as other methods for generating Illumina next-generation sequencing library preparations, such as described in Oyola et al. (2012). In other embodiments, SOLiD TM or Ion Torrent sequencing compatible methods (e.g., Fragment library construction kit, Mate-Paired library construction kit, ChIP-Seq kit, Total RNA-Seq Kit, SAGE TM Reagent test kit, The nucleic acid library is generated using a RNA-Seq library construction kit, etc. Other methods for next generation sequencing methods, such as those described in Pareek (2011) and Thudi (2012), include various methods for library construction that can be used with embodiments of the present invention.

[0151] In certain aspects, the sequencing technology used in the methods of the present disclosure includes HiSeq from Illumina, Inc. TM System (e.g., HiSeq TM 2000 and HiSeq TM 1000)NextSeq TM 500 System and MiSeq TM System. HiSeq TMThe system is based on massively parallel sequencing of millions of fragments, using ligation of randomly fragmented genomic DNA to flat surfaces, optically transparent surfaces, and solid-phase amplification to create a high-density sequencing flow cell with millions of clusters, each containing approximately 1,000 copies of template per square centimeter. These templates are sequenced using four-color DNA sequencing by synthesis technology. MiSeq TM System using TruSeq TM , Illumina sequencing by synthesis based on reversible terminators.

[0152] Another example of a DNA sequencing technique that can be used in the methods of the present disclosure is 454 sequencing (Roche) (Margulies et al., 2005). 454 sequencing involves two steps. In the first step, the DNA is sheared into fragments of approximately 300-800 base pairs, and the fragments are then blunt-ended. Oligonucleotide adapters are then attached to the ends of the fragments. The adapters are used as primers for amplification and sequencing of the fragments. The fragments can be attached to DNA capture beads, such as streptavidin-coated beads, using, for example, adapter B containing a 5′-biotin tag. In droplets of an oil-water emulsion, PCR amplifies the fragments attached to the beads. The result is that there are multiple copies of the cloned amplified DNA fragments on each bead. In the second step, the beads are captured in a hole (picoliter size). Pyrophosphate sequencing is performed in parallel for each DNA fragment. The addition of one or more nucleotides generates a light signal, which is recorded by a CCD camera in a sequencing instrument. The signal intensity is proportional to the number of nucleotides incorporated.

[0153] Another example of a DNA sequencing technology that can be used in the methods of the present disclosure is SOLiD technology (Life Technologies, Inc.). In SOLiD sequencing, genomic DNA is sheared into fragments, and adapters are connected to the 5' and 3' ends of the fragments to generate a fragment library. Alternatively, internal adapters can be introduced by connecting adapters to the 5' and 3' ends of the fragments, circularizing the fragments, digesting the circularized fragments to produce internal adapters, and connecting the adapters to the 5' and 3' ends of the resulting fragments to generate a mate-paired library. Next, a clonal bead population is prepared in a microreactor containing beads, primers, templates, and PCR components. After PCR, the template is denatured and the beads are enriched to separate beads with extended templates. The template on the selected beads is 3' modified so that it can be bound to a glass slide.

[0154] Another example of DNA sequencing technology that can be used in the methods of the present disclosure is the IonTorrent system (Life Technologies, Inc.). The Ion Torrent uses a high-density array of micro-machined wells to perform this biochemical process in a massively parallel manner. Each well holds a different DNA template. Below the well is an ion-sensitive layer, and below the ion-sensitive layer is a proprietary ion sensor. If a nucleotide (such as C) is added to the DNA template and then incorporated into the DNA chain, a hydrogen ion is released. The charge from this ion changes the pH of the solution, which can be detected by a proprietary ion sensor. The sequencer will call the base, converting it directly from chemical information to digital information. The Ion Personal Genome Machine (PGM) then TM ) The sequencer sequentially floods the chip with nucleotides one by one. If the next nucleotide flooding the chip does not match, no voltage change is recorded and no base is called. If there are two identical bases on the DNA strand, the voltage will double and the chip will record two identical bases being called. Because this is direct detection, there is no scanning, no camera, and no light, so the incorporation of each nucleotide is recorded in seconds.

[0155] Another example of a sequencing technology that can be used in the methods of the present disclosure includes the Single Molecule Real-Time (SMRT) sequencing technology from Pacific Biosciences. TM ) technology. TM In the method, each of the four DNA bases is connected to one of four different fluorescent dyes. These dyes are phosphate-linked. A single DNA polymerase is fixed using a single-molecule template single-stranded DNA at the bottom of a zero-mode waveguide (ZMW). ZMW is a confined structure that enables the observation of the incorporation of a single nucleotide by DNA polymerase (in microseconds) against the background of fluorescent nucleotides that rapidly diffuse into and diffuse out of the ZMW. It takes several milliseconds to incorporate a nucleotide into a growing chain. During this time, the fluorescent label is excited and produces a fluorescent signal, and the fluorescent label is cut off. The detection of the corresponding fluorescence of the dye indicates which base has been incorporated. Repeat the process.

[0156] Another sequencing platform includes the CGA platform (complete genomics). The CGA technology is based on the preparation of circular DNA libraries and rolling circle amplification (RCA) to produce DNA nanoballs arranged on a solid support (Drmanac et al. 2009). The CGA platform of complete genomics uses a novel strategy called combined probe anchor ligation (cPAL) for sequencing. The process begins with hybridization between an anchor molecule and one of the only adapters. Four degenerate 9-mer oligonucleotides are labeled with specific fluorophores corresponding to specific nucleotides (A, C, G or T) at the first position of the probe. Sequencing occurs in a reaction in which the correct matching probe hybridizes with the template and is connected to the anchor using T4 DNA ligase. After imaging the ligation product, the connected anchor probe molecule is denatured. The hybridization, connection, imaging and denaturation process are repeated five times using a new fluorescently labeled 9-mer probe set containing known bases at positions n+1, n+2, n+3 and n+4.

[0157] XI. Kit

[0158] The technology herein includes a kit for analyzing copy number variation or allele frequency in a DNA sample. "Kit" refers to a combination of physical elements. For example, a kit may include, for example, one or more components, such as nucleic acid primers, enzymes, reaction buffers, instructions, and other elements that can be used to implement the technology described herein. These physical elements can be arranged in any manner suitable for performing the present invention.

[0159] The components of the test kit can be packaged in an aqueous medium or in a lyophilized form. The container device of the test kit will generally include at least one vial, test tube, flask, bottle, syringe or other container device, the component can be placed therein, and preferably appropriately divided and placed (for example, divided into the wells of a microtiter plate). If there are multiple components in the test kit, the test kit will generally also contain a second, third or other additional container, and the additional components can be placed therein separately. However, various combinations of components can be included in a single vial. The test kit of the present invention will generally also include a device for accommodating nucleic acid, and any other reagent container for commercial sale in a sealed manner. Such a container can include a plastic container for injection or blow molding in which the desired vial is retained. The test kit will also include instructions for the use of any other reagent not included in the test kit components and the test kit. The instructions can include variants that can be realized.

[0160] XII. Examples

[0161] The following examples are included to illustrate preferred embodiments of the present invention. It will be appreciated by those skilled in the art that the techniques disclosed in the following examples represent techniques that the inventors have found to work well in the practice of the present invention and therefore can be considered to constitute preferred modes for the practice of the present invention. However, based on this disclosure, it will be appreciated by those skilled in the art that many changes may be made to the specific embodiments disclosed and still the same or similar results may be obtained without departing from the spirit and scope of the present invention.

[0162] Example 1 - Calibration Results

[0163] An exemplary calibration experiment for the ERBB2 QASeq panel was performed on a normal cell line gDNA sample NA18562, which should not contain ERBB2 amplification, to analyze quantitation variability and potential LoD. The workflow is as described in the "QASeq Workflow" section. Taq polymerase is used for all PCR steps. Denaturation is performed at 95°C and annealing / extension is performed at 60°C (except for the universal PCR step, where annealing / extension is performed at 68°C). Because all original molecules with UMIs attached need to appear in the NGS output, 15 reads are retained for each molecule / UMI. For an input of 2500 haploid genome copies and 20 amplicon combinations, the total reads required are approximately 2×2500×20×15=1,500,000. Note that in this workflow, each strand in a DNA duplex carries a different UMI, so 2500 haploid genome copies=5000 molecules=8.3ng gDNA. The experiments were performed on an Illumina MiSeq instrument.

[0164] NGS reads were aligned to amplicon sequences using exact string matching; the alignment rate for different libraries ranged from 50% to 70%. Next, UMI family size and number of unique UMIs were analyzed. For most loci, the distribution of UMI family size peaked at ≈20 ( Figure 5 ). Remove UMI families containing obvious PCR errors (i.e., G bases found in poly(H)UMI sequences) and UMIs with a family size of less than 4 ( Figure 5 ). If the UMI mapping rate is perfect, the number of unique UMIs should be equal to the original number of molecules in the sample. For an input of 2500 haploid genome copies (5000 molecules), the number of unique UMIs obtained ranged from 632 to 3065 depending on the locus ( Figure 6 ).

[0165] To estimate the LoD of the assay, libraries were prepared for four different DNA inputs: 75, 250, 750, and 2500 haploid genome copies; each condition was repeated five times. The CNV ratios of the samples were calculated as described in the “Data Analysis Workflow” section. The CNV ratios (σ CNV比率 ) to assess quantitative variability; the LoD of the assay can be estimated as 3σ CNV比率 Simulations were also performed to calculate the theoretical σ CNV比率 ; Note that if the number of input molecules increases, σ CNV比率 and LoD should be reduced. σ CNV比率 Higher than the theoretical value ( Figure 7 ), which is as expected since UMI connection bias and amplification bias cannot be eliminated. For 2500 haploid genome copies input, the current best σ CNV比率 1%; to be conservative, a linear approximation based on all 4 data points was used and σ CNV比率 = 2%; therefore, the estimated LoD is approximately 6% of the extra copies. Based on extrapolation to an input of 50,000 haploid genome copies, the potential σ CNV比率 is 0.3%, and the LoD is about 1%. Another way to estimate the LoD is to test a series of calibration samples containing extra copies at different frequencies; the lowest detectable frequency of the extra copies is the LoD.

[0166] Example 2 - CNV detection results in FFPE samples

[0167] Two FFPE slides were analyzed using the ERBB2 panel example described in the "Multiplex PCR Panel Design" section and Example 1. The FFPE slides (purchased from Asterand) were from the same lung cancer tumor and were not expected to contain ERBB2 CNV. First, DNA was extracted using the QIAamp DNA FFPE Tissue Kit (Qiagen), and >6 μg of DNA was obtained for each sample. Libraries were prepared using the same method as described in Example 1. Each library used 8.3 ng of extracted DNA, equivalent to 2500 haploid genome copies and 5000 molecular inputs. The number of NGS reads retained for each library (1,500,000 reads) was the same as the 2500 haploid genome copy number input cell line gDNA library.

[0168] Data analysis was performed using the same methods as described in Example 1. A similar UMI family size distribution pattern was obtained as for the cell line gDNA library (Figure 8A). The number of unique UMIs was less than that of the input cell line gDNA library with 2500 haploid genome copies. The UMI connection rate of the FFPE sample was on average about 1 / 4 of that of the cell line gDNA, indicating that more than 300% of FFPE DNA needs to be loaded to achieve the same LoD as the cell line gDNA sample (Figure 8B).

[0169] The calculated CNV ratios for the FFPE samples are shown in Figure 8C. The inferred LoD of this assay = 15% is based on the calibration results of 750 haploid genomic copies input cell line gDNA, which has a similar number of unique UMIs as the FFPE library. Based on the current results, no CNVs of ERBB2 were detected in these FFPE slides. Since the LoD decreases with increasing input molecule number, a LoD of 6% can be achieved based on the calibration results of 2500 haploid genomic copies input cell line gDNA.

[0170] Example 3 - CNV Quantification Results Spiked into Clinical FFPE Samples

[0171] A 100-plex QASeq panel was used to quantify the ploidy of ERBB2 in breast cancer FFPE samples. A 50-plex was located in the ERBB2 gene region (primer sequences are shown in Table 3; primer names have "ERBB2"), and a 50-plex was located in the short arm of chromosome 17 as a reference (primer sequences are shown in Table 3; primer names have "Ref").

[0172] Two previously characterized FFPE DNA samples (1 "normal" sample and 1 "ERBB2 amplification abnormal" sample) were mixed to generate 2.5%, 5%, and 10% ERBB2 FEC samples. The "normal" sample DNA was extracted from a FFPE lung cancer sample (purchased from Asterand), which should not have ERBB2 amplification (FEC = 0%); the "ERBB2 amplification abnormal" sample DNA was extracted from a FFPE breast cancer sample (purchased from OriGene), which has an ERBB2 FEC of 78%. The sample input for each library was 8.3ng DNA (quantified by qPCR). The "normal" sample was tested using 5 separately prepared replicate NGS libraries, each with 8.3ng DNA input. The experimental normalized FEC values ​​are shown in Fig.13 .Normalized FEC is calculated as follows:

[0173] Normalized FEC 样品 =(1+FEC 样品 ) / (1+FEC 正常样品 )-1

[0174] FEC 正常样品 is the average of 5 replicates. The LoD of the CNV combination is estimated as:

[0175] FEC LoD =3×σ 正常样品 / (1+FEC 正常样品 )=0.85%

[0176] Here, σ 正常样品 is the standard deviation of 5 replicates. CNVs were successfully detected in 2.5%, 5%, and 10% of the ERBB2 FEC samples because their calculated FECs were outside the 3 standard deviation range (see Fig.13 ). The experimental normalized FEC of ERBB2 was closely related to the expected value.

[0177] Example 4 - Comprehensive Panel for Mutation and CNV Quantification

[0178] The proposed method (QASeq) can be used not only for CNV quantification, but also for NGS error correction and mutation quantification. In each QASeq amplicon, the region between the 3′ of fP and the 3′ of rPin is the mutation detection region (MDR); any small variation in the MDR (including base substitutions, deletions, and insertions less than 500 bp) can be detected with an LoD of 0.1%-0.3%. This is much better than the standard non-UMI NGS method for mutation detection, which has an LoD ≈ 1%.

[0179] A 179-plex comprehensive panel was developed and tested for mutation and CNV quantification in breast cancer samples. Each plex contains 3 primers: fP (also known as SfP), rPin (also known as SrPB) and rPout (also known as SrPA), as described in the previous section. 95 primer sets were used alone for CNV quantification, 45 of which were used for the gene ERBB2 and 50 for the short arm of chromosome 17 as a reference. Five primer sets for the ERBB2 gene were used for CNV and mutation quantification. Another 79 primer sets were used only for mutation quantification. UfP and UrP were used for universal amplification (sequences are shown in Table 3).

[0180] The method for CNV quantification is the same as in the previous section; the data processing workflow for mutation quantification is summarized in Fig.14.After optional adapter trimming, the NGS reads are aligned to the amplicon sequences. At each locus, the reads are grouped into UMI families; UMI families with errors in the UMI sequence are removed, and UMI families with small UMI family size (≤3) are also removed. Next, the consensus MDR sequence is found for each UMI family, usually the MDR sequence with the most occurrences in the UMI family. The last step is to compare the consensus sequence with the wild-type MDR sequence and perform de novo mutation calling. The VAF of a mutation can be calculated as: VAF = number of UMI families with the mutation / total number of UMI families.

[0181] The 179-recombination was tested based on the Multiplex I cfDNA Reference Standard Set from Horizon Discovery. Three replicate NGS libraries of the Wild Type cfDNA Reference Standard and three replicates of the 0.3% cfDNA reference standard (created by mixing 0.1% cfDNA Reference Standard and 1% cfDNA Reference Standard) were tested. The sample input for each library was 8.3 ng DNA (quantified by qPCR).

[0182] The overall on-target rate for all libraries was greater than 50% (i.e., >50% of NGS reads could be aligned to amplicons); the conversion rate (i.e., the percentage of input molecules sequenced) averaged 62%, and 97% of the panel had a conversion rate >10% (see Fig.15 ). The error rates after UMI correction varied at different nucleotide positions; in the three replicate libraries of the Horizon Discovery Multiplex I Wild Type cfDNA Reference Standard, the highest error rates were 0.23%, 0.20%, and 0.23%, and the average error rates were 0.006%, 0.005%, and 0.005% (see Fig.16 ). The ability to quantify mutations was validated using the 0.3% cfDNA Reference Standard. The experimental VAFs for the six mutations were generally consistent with the expected VAFs; the differences were mainly due to the randomness of sampling a small number (≤9) of mutant molecules (see Fig.17 ).

[0183] ***

[0184] In view of the present disclosure, all methods disclosed and claimed herein can be performed and executed without undue experimentation. Although the compositions and methods of the present invention have been described in accordance with preferred embodiments, it will be apparent to those skilled in the art that changes may be applied to the methods described herein and to the steps of the methods described herein or to the sequence of steps of the methods described herein without departing from the concept, spirit and scope of the present invention. More specifically, it will be apparent that certain agents that are both chemically and physiologically related can replace the agents described herein while obtaining the same or similar results. All such similar substitutions and modifications apparent to those skilled in the art are considered to fall within the spirit, scope and concept of the present invention as defined by the appended claims.

[0185] References

[0186] To the extent that the following references provide exemplary procedures or other details supplementary to those described herein, the references are expressly incorporated herein by reference.

[0187] Lun et al., "NoinvasiVe prenatal diagnosis of monogenic diseases by digital size selection and relatiVe mutation dosage on DNA in maternal plasma," Proc. Natl. Acad. Sci. USA, 105: 19920-19925, 2008.

Claims

1. A method for preparing a targeted region of genomic DNA for high-throughput sequencing, the method comprising: (a) obtaining a genomic DNA sample; (b) amplifying at least a portion of the genomic DNA sample by performing two cycles of PCR using: (i) a first oligonucleotide comprising, from 5′ to 3′, a first region, a second region as a spacer having a length of 1 to 50 nucleotides, a third region as a unique molecular identifier (UMI) comprising at least four degenerate nucleotides, and a fourth region comprising a sequence complementary to a first target genomic DNA region of a target genomic locus; and (ii) a second oligonucleotide comprising, from 5′ to 3′, a fifth region, a sixth region as a spacer having a length of 1 to 50 nucleotides, and a seventh region comprising a sequence complementary to a second target genomic DNA region of the target genomic locus, The melting temperatures of the first region and the fifth region are 0-10° C. higher than the melting temperatures of the fourth region and the seventh region; (c) amplifying the product of step (b) using an annealing temperature that is 0-10° C. higher than the annealing temperature used in step (b) and performing at least three PCR cycles using the following: (i) a third oligonucleotide comprising a sequence capable of hybridizing to a reverse complementary sequence of at least a portion of the first region; and (ii) a fourth oligonucleotide comprising a sequence capable of hybridizing to a reverse complementary sequence of at least a portion of the fifth region; and (d) amplifying the product of step (c) by performing at least one PCR cycle using a fifth oligonucleotide, wherein the fifth oligonucleotide includes, from 5' to 3', an eighth region comprising at least a portion of a next generation sequencing (NGS) adapter, a ninth region having a length of 1 to 50 nucleotides as a spacer, and a tenth region comprising a sequence complementary to a third target genomic DNA region of the target genomic locus, wherein the third target genomic DNA region is closer to the first target genomic DNA region by at least one nucleotide than the second target genomic DNA region is to the first target genomic DNA region. 2 . The method according to claim 1 , wherein the method is a method for preparing 1 to 10,000 targeted regions of genomic DNA for high-throughput sequencing.

3. The method according to any one of claims 1-2, wherein the third target genomic DNA region is 1-10 bases closer to the first target genomic DNA region than the second target genomic DNA region is to the first target genomic DNA region.

4. The method according to any one of claims 1 to 3, wherein the first region and the eighth region are universal primer binding sites.

5. The method according to any one of claims 1-4, wherein the first region and the eighth region comprise complete NGS adapter sequences.

6. The method according to any one of claims 1-5, wherein the fifth region comprises a sequence that is not found in the human genome.

7. The method of any one of claims 1-6, wherein the fifth region comprises a sequence different from an NGS adapter sequence.

8. The method according to any one of claims 1 to 7, wherein the degenerate nucleotides in the third region are each independently one of A, T or C.

9. The method according to any one of claims 1 to 8, wherein none of the degenerate nucleotides in the third region is G.

10. The method of any one of claims 1-9, wherein there is a population of first oligonucleotides each having a unique third region.

11. The method according to any one of claims 1 to 10, further comprising purifying the product of step (c).

12. The method of claim 11, wherein the purification comprises SPRI purification or column purification.

13. The method according to any one of claims 1 to 12, further comprising purifying the product of step (d).

14. The method of claim 13, wherein purification comprises SPRI purification or column purification.

15. The method according to any one of claims 1 to 14, further comprising: (e) amplifying the product of step (d) by PCR using primers that hybridize to the first region and the eighth region, wherein the primers include an index sequence for next generation sequencing.

16. The method of claim 15, further comprising purifying the product of step (e).

17. The method of claim 16, wherein purification comprises SPRI purification or column purification.

18. The method according to any one of claims 15 to 17, further comprising: (f) performing high-throughput DNA sequencing on the product of step (e).

19. The method of claim 18, wherein high-throughput DNA sequencing comprises next generation sequencing.

20. The method of any one of claims 1-19, wherein the first target genomic DNA region and the second target genomic DNA region are on opposite strands of genomic DNA.

21. The method of any one of claims 1-20, wherein the first target genomic DNA region and the second target genomic DNA region are separated by 40 nucleotides to 500 nucleotides.

22. The method of any one of claims 1-21, wherein step (b) comprises an extension time of about 30 minutes.

23. The method of any one of claims 1-22, wherein step (c) comprises an extension time of about 30 seconds.

24. The method of any one of claims 1-23, wherein step (d) comprises an extension time of about 30 minutes.

25. A method for quantifying the frequency of extra copies (FEC) of at least one target gene, the method comprising: (a) obtaining a genomic DNA sample; (b) preparing the genomic DNA for high-throughput sequencing according to the method according to any one of claims 1 to 24, wherein the sequences of the fourth region, the seventh region, and the tenth region hybridize with the at least one target gene; (c) performing high-throughput sequencing according to the method of claim 18; and (d) calculating the FEC of the at least one target gene based on the sequencing information obtained in step (c).

26. The method according to claim 25, wherein the method is a method for quantifying the FEC of a set of target genes, wherein the set of target genes includes 2 to 1000 target genes.

27. The method of claim 25 or 26, wherein step (b) is performed using a first oligonucleotide population, a second oligonucleotide population, and a fifth oligonucleotide population, wherein a portion of each of the first, second, and fifth oligonucleotide populations comprises a fourth, seventh, and tenth region complementary to one of the set of target genes, respectively.

28. The method of any one of claims 25-27, wherein each of the fourth, seventh and tenth regions comprises a sequence that is found only once in the human genome.

29. The method of any one of claims 25-28, wherein each first oligonucleotide that hybridizes to a target gene has a unique third region compared to each other first oligonucleotide that hybridizes to the same target gene.

30. The method according to any one of claims 25 to 29, wherein step (b) is performed using a first oligonucleotide, a second oligonucleotide and a fifth oligonucleotide comprising a fourth, a seventh and a tenth region respectively complementary to a reference gene.

31. The method according to any one of claims 25-30, wherein step (b) prepares a portion of each target gene or reference gene for high-throughput sequencing, wherein the length of the portion is between 40 nucleotides and 500 nucleotides.

32. The method according to any one of claims 25-31, wherein FEC is defined as follows:

33. The method of any one of claims 25-32, wherein step (d) comprises: (i) aligning the NGS reads to the targeted portion of each target gene and grouping the NGS reads into subsets based on the loci to which they align; (ii) partitioning the NGS reads based on their UMI sequences at each locus so that all NGS reads carrying the same UMI sequence are grouped into a UMI family; (iii) removing UMI families obtained due to PCR errors or NGS errors; (iv) calculating the number of unique UMI sequences for each locus; and (v) calculating the FEC based on the number of unique UMI sequences for each locus in each target gene and reference gene.

34. The method of claim 33, wherein step (d)(iii) comprises removing UMI sequences that do not satisfy the UMI degenerate base design.

35. The method of claim 33 or 34, wherein step (d)(iii) comprises removing UMI families whose UMI family size is less than Fmin, wherein the UMI family size is the number of reads carrying the same UMI, wherein Fmin is between 2 and 20.

36. The method of any one of claims 33-35, wherein step (d)(iv) comprises removing UMI sequences that differ by only one or two bases from another UMI sequence with a larger family size.

37. The method according to any one of claims 25-36, wherein the FEC is defined as follows: in is the sum of the number of unique UMIs of all or part of the target gene loci, u is the number of loci to be considered, and u does not exceed the total number of loci in the target gene; is the sum of the number of unique UMIs of all or part of the reference loci, v is the number of loci to be considered for a reference, v does not exceed the total number of loci in the reference; w is the number of references to be considered, w does not exceed the total number of references; and k is determined by experimental calibration.

38. The method of any one of claims 25-37, wherein the FEC is used to identify the copy number variation (CNV) status of the target gene.

39. A method for quantifying allele ratios of different genetic identities of at least one target genomic locus, the method comprising: (a) obtaining a genomic DNA sample; (b) preparing the genomic DNA for high-throughput sequencing according to the method of any one of claims 1 to 24, wherein the sequences of the fourth region, the seventh region, and the tenth region hybridize with the genomic DNA near the at least one target genomic locus; (c) performing high-throughput sequencing according to the method of claim 18; and (d) calculating the allele ratios of different genetic identities of the at least one target genomic locus based on the sequencing information obtained in step (c).

40. The method of claim 39, wherein the method is a method for quantifying the allele ratios of different genetic identities of a set of target genomic loci, wherein the set of target genomic loci comprises 2 to 10,000 target genomic loci.

41. A method according to claim 39 or 40, wherein step (b) is performed using a first oligonucleotide population, a second oligonucleotide population, and a fifth oligonucleotide population, wherein a portion of each of the first, second, and fifth oligonucleotide populations respectively comprises a fourth, seventh, and tenth region complementary to genomic DNA near at least one of the set of target genomic loci.

42. The method of any one of claims 39-41, wherein each of the fourth, seventh and tenth regions comprises a sequence that is not hybridizable to a non-target region of the genomic DNA under the conditions of step (b).

43. A method according to any one of claims 39-42, wherein each first oligonucleotide that hybridizes to the genomic DNA near a target genomic locus has a unique third region compared to each other first oligonucleotide that hybridizes to the genomic DNA near the same target genomic locus.

44. The method of any one of claims 39-43, wherein each target genomic locus is between 40 nucleotides and 500 nucleotides in length.

45. The method of any one of claims 39-44, wherein step (d) comprises: (i) aligning NGS reads to targeted genomic loci and grouping the NGS reads into subgroups based on the loci to which they align; (ii) partitioning the NGS reads based on their UMI sequences at each locus so that all NGS reads carrying the same UMI sequence are grouped into a UMI family; (iii) removing UMI families obtained due to PCR errors or NGS errors; (iv) calling the genetic identity for each remaining UMI family; (v) calculating the number of unique UMI sequences for each locus; and (vi) calculating the allele ratio.

46. ​​The method of claim 45, wherein step (d)(iii) comprises removing UMI sequences that do not satisfy the UMI degenerate base design.

47. The method of claim 45 or 46, wherein step (d)(iii) comprises removing UMI families whose UMI family size is less than Fmin, wherein the UMI family size is the number of reads carrying the same UMI, wherein Fmin is between 2 and 20.

48. The method of any one of claims 45-47, wherein step (d)(iii) comprises removing UMI sequences that differ by only one or two bases from another UMI sequence with a larger family size.

49. The method of any one of claims 45-48, wherein step (d)(iv) comprises calling the genetic identity only when at least 70% of the reads in a UMI family are identical at the target genetic locus.

50. The method according to any one of claims 39-49, wherein the allele ratio is defined as R 等位基因 =N1 / N2, where N1 is the number of unique UMIs of the first genetic identity and N2 is the number of unique UMIs of the second genetic identity.

51. The method of any one of claims 45-49, wherein step (d)(iv) comprises identifying a consensus sequence for each UMI family.

52. The method of claim 51, wherein the consensus sequence is a sequence that occurs most frequently in the UMI family.

53. The method of claim 51 or 52, further comprising comparing the consensus sequence to a wild-type sequence of the locus, thereby identifying a mutation in the consensus sequence.

54. The method of claim 53, further comprising calculating the variant allele frequency (VAF) of the identified mutation.

55. The method of claim 54, wherein the VAF of an identified mutation is defined as the number of UMI families with the mutation / the total number of UMI families.

Images