Methods and apparatus for targeted region tandem repeat variation typing based on long read sequencing

Abstract

The application provides a target region tandem repeat variation typing method and device based on long read sequencing. The method comprises the following steps: obtaining long read sequencing data of a target region, screening sequencing reads located in the target region to obtain target sequencing reads; identifying the sequence of a tandem repeat unit in the target region to obtain the position and copy number of the tandem repeat unit; detecting multi-source variation information in the target region to obtain multi-source variation data; performing haplotype attribution determination on each target sequencing read to obtain a preliminary haplotype typing result; and based on the proportion distribution and / or copy number distribution characteristics of target sequencing reads in the first haplotype, the second haplotype and the unassigned category, reassigning the target sequencing reads in the unassigned category to obtain a final haplotype typing result. The application realizes high-precision, high-resolution haplotype-specific typing and visual display.

Description

Technical Field

[0001] This invention relates to the field of gene sequencing technology, and more specifically, to a method and apparatus for typing target region tandem repeat variants based on long-read sequencing. Background Technology

[0002] Tandem repeats (TRs) are short DNA sequences arranged in a continuous repeating pattern within a specific region of the genome. Based on the length of the repeating unit, tandem repeats can generally be divided into two categories: short tandem repeats (STRs), with repeating units typically ranging from 2 to 6 bp, and variable-number tandem repeats (VNTRs), with repeating units typically ranging from 7 to 100 bp. These repeating sequences play important roles in genome regulation, structure, and evolution. Due to their high variability, they are also widely used in genetic research (such as Huntington's disease and Fragile X syndrome), forensic individual identification, and population genetic analysis.

[0003] Traditional next-generation sequencing (NGS) (such as the Illumina platform) often struggles to cover long or complex TR regions due to limited read lengths (typically <300bp), resulting in uneven coverage, high alignment error rates, and consequently, copy number estimation bias and genotyping errors. With the development of long-read sequencing technologies (third-generation sequencing, TGS), especially the widespread adoption of PacBio SingleMolecule Real-Time (SMRT) sequencing and Oxford Nanopore Technologies (ONT) sequencing platforms, single-read reads can reach tens or even hundreds of kb, directly traversing the entire TR region. This demonstrates significant advantages in TR detection accuracy and genotyping consistency.

[0004] Currently, TR detection software based on TGS data has been developed to some extent, but each has its own focus and limitations:

[0005] Straglr primarily uses a parameter-free mode and is designed to detect newly emerging or significantly amplified TRs in the genome. It does not rely on a reference TR library, making it difficult to perform precise quantitative analysis and typing of specific known pathogenic sites.

[0006] TriColor: Also uses a parameter-free mode, suitable for the discovery of new TRs across the entire genome, but has limited quantitative accuracy for specific known sites, and the output results are relatively macroscopic, lacking detailed analysis at the single-molecule level.

[0007] TRGT: Detects based on known TR regions, supports clustering and length-based genotyping, and can analyze methylation status. However, TRGT primarily provides overall genotyping results for each sample and cannot directly output haplotype attribution information for each sequencing read. When two haplotypes in a sample are highly similar in length or sequence, the genotyping accuracy of TRGT decreases significantly.

[0008] Other tools, such as RepeatHMM-Long, TredPARSE, and Tandem-genotypes, have certain advantages in specific site detection or whole-genome scanning, but they still have problems such as misclassification and inaccurate read allocation in long repeats, complex repeats, or extremely high sequencing depths.

[0009] Although third-generation sequencing technology has significantly improved TR detection capabilities, existing technologies still have significant gaps and deficiencies in applications involving high-depth targeted PCR product sequencing data. Specifically:

[0010] 1) Lack of precise read-by-read structural analysis: Most mainstream software assumes moderate coverage depth and performs genotyping statistics on a sample-by-sample basis, failing to directly analyze the copy number (CPN) and sequence structure of high-depth targeted sequencing data generated after PCR enrichment. In high-depth sequencing data, each read carries independent molecular information; ignoring single-molecule-level analysis can lead to missed detections of rare alleles or complex structural variations in heterozygotes. Furthermore, for disease samples, studying the breaks between repetitive sequences (e.g., the break of GA in CTG repeats, such as CTGCTGCTGGACTG) is also significant. Studies have shown that the break characteristics differ across diseases.

[0011] 2) Insufficient Haplotyping Capability: In diploid organisms, distinguishing between two alleles (i.e., haplotypes) from the paternal and maternal parents is crucial. While existing tools like TRGT provide genotyping results, they often cannot trace haplotype information back to specific sequencing reads. When two haplotypes have similar TR lengths or high sequence similarity, existing algorithms struggle to accurately differentiate them, leading to genotyping errors. Furthermore, single-molecule-level SNP / Indel-assisted genotyping tools (such as HiPhase) typically require specialized variant detection workflows and lack integrated solutions specifically for TR regions.

[0012] 3) Lack of genotyping correction mechanisms in high-depth data: In high-depth sequencing of PCR products, amplification bias or sequencing errors may lead to an imbalance in the proportion of reads for certain haplotypes or the appearance of unassigned reads. Current technologies lack effective statistical correction algorithms to handle these marginal cases (such as misclassification of homozygotes, severe imbalance in haplotype ratios, etc.), limiting their accuracy in clinical diagnosis.

[0013] Therefore, there is an urgent need to develop a detection method and system that can accurately identify tandem repeat copy number at the single read level based on third-generation sequencing data and high-depth sequencing results (e.g., high-depth sequencing results of PCR products), perform haplotype typing by combining multi-source variant information, and have an intelligent correction mechanism. Summary of the Invention

[0014] The main objective of this invention is to provide a genotyping method for tandem repeat variants in target regions based on long-read sequencing, in order to solve the problems of inaccurate genotyping of targeted high-depth sequencing data, lack of single-molecule structure visualization, and inability to accurately classify haplotypes in the prior art.

[0015] To achieve the above objectives, according to a first aspect of the present invention, a method for genotyping target region tandem repeat variants based on long-read sequencing is provided, the method comprising:

[0016] S1. Obtain long-read sequencing data of the target region, align the long-read sequencing data to the reference genome, generate an alignment file, and screen out the sequencing reads located in the target region to obtain the target sequencing reads.

[0017] S2, perform sequence parsing on each of the above target sequencing reads, identify the sequence of the tandem repeat unit in the above target region, and obtain the position and copy number of the above tandem repeat unit;

[0018] S3. Based on the above comparison file, detect multi-source variation information in the above target area to obtain multi-source variation data;

[0019] S4. Using the haplotype construction algorithm, combined with the above copy number and the above multi-source variant data, the haplotype attribution of each of the above target sequencing reads is determined, and each of the above target sequencing reads is classified into the first haplotype, the second haplotype, or the unassigned category according to the attribution, so as to obtain the preliminary haplotype typing results.

[0020] S5. Based on the proportion and / or copy number distribution characteristics of the target sequencing reads in the first haplotype, the second haplotype and the unassigned categories mentioned above, the target sequencing reads in the unassigned categories are redistributed to obtain the final haplotype typing result.

[0021] Among them, the above-mentioned multi-source variation information includes any of the following: single nucleotide variation, small fragment insertion / deletion and structural variation; the above-mentioned tandem repeat unit is one or more.

[0022] Furthermore, S2 above includes the following steps:

[0023] S21, Construct the regular expression or sequence pattern matching rule for the above-mentioned serial repeating units;

[0024] S22, use the above regular expression or sequence pattern matching rules to scan the target sequencing reads and locate the positions of all matched tandem repeat units.

[0025] S23, count the number of tandem repeat units matched in each of the above target sequencing reads to obtain the above copy number.

[0026] Furthermore, in S1 above, when the long-read sequencing data is sequencing data of PCR products, after screening out the sequencing reads located in the target region and before obtaining the target sequencing reads, it is necessary to verify the validity of the sequencing reads based on primer sequence matching degree, retain the sequencing reads that have passed the sequencing test, and obtain the target sequencing reads from the sequencing reads that have passed the sequencing test.

[0027] Furthermore, the steps for obtaining the target sequencing reads from the above-mentioned sequencing reads include: removing the primer sequences and flanking sequences at both ends of the above-mentioned sequencing reads to obtain truncated reads, which are the target sequencing reads.

[0028] Furthermore, S3 above includes the following steps:

[0029] S31, Use a single nucleotide variant detection tool to analyze the above alignment file and extract single nucleotide variant site information and small fragment insertion / deletion site information in the target region;

[0030] S32, Use a structural variation detection tool to analyze the above-mentioned comparison files and extract structural variation site information within the target region;

[0031] S33. Integrate the above single nucleotide variant site information, small fragment insertion / deletion site information, and above structural variant site information to obtain the above multi-source variant data.

[0032] Furthermore, the single nucleotide variant detection tools mentioned above are DeepVariant or longshot, and the structural variant detection tools mentioned above are Sawfish, cuteSV, or clair3.

[0033] Furthermore, S4 above includes the following steps:

[0034] S41, merge the above copy number and the above multi-source mutation data to generate a comprehensive mutation VCF file;

[0035] S42, input the above-mentioned integrated variant VCF file into the above-mentioned haplotype construction algorithm, perform haplotype phase inference on each of the above-mentioned target sequencing reads, and output the haplotype category to which each target sequencing read belongs.

[0036] Furthermore, S5 above includes the following steps:

[0037] S51, count the number of target sequencing reads in the first haplotype, the second haplotype and the unassigned category, and calculate the proportion of each to the total valid sequencing reads;

[0038] S52, determine whether the proportions of the first haplotype and the second haplotype are both greater than or equal to the first preset threshold; if so, determine that the typing result is stable, do not redistribute, and take the current typing result as the final typing result of the haplotype; if not, execute S53;

[0039] S53, determine that the proportion of the first haplotype or the second haplotype is greater than or equal to a second preset threshold, and the proportion of the unassigned category is greater than or equal to a third preset threshold; if so, recalibrate the unassigned category to the haplotype category with the smaller proportion among the first haplotype and the second haplotype, and recalibrate the haplotype category with the smaller proportion to the unassigned category to obtain a new classification result, and record the new classification result as the final haplotype classification result; if not, execute S54;

[0040] S54, obtain the histogram of copy number distribution of tandem repeat units in all the above target sequencing reads, identify the copy number intervals corresponding to the first two highest peaks, and use the midpoint of these two copy number intervals as the copy number threshold.

[0041] S55. Based on the above copy number threshold, classify all the above target sequencing reads into two haplotypes; calculate the ratio of the two haplotypes and determine whether the ratio of the two haplotypes is (4~10):1. If so, the two haplotypes are the final haplotype typing result. If not, proceed to S56.

[0042] S56, Randomly assign the above target sequencing reads to two haplotypes in a 1:1 ratio, and the above two haplotypes are the final haplotype typing results;

[0043] The first, second, and third preset thresholds are all 30% to 35%.

[0044] Furthermore, after obtaining the final haplotype typing results, the target region tandem repeat variant typing method further includes:

[0045] S6. Based on the final haplotype typing results above, for each haplotype, a consensus sequence is generated based on the tandem repeat unit sequences of the target sequencing reads contained therein. The copy number statistics and sequence purity of each haplotype are calculated, and visualization results are generated.

[0046] Furthermore, S6 above includes the following steps:

[0047] S61, extract the sequence of the tandem repeat unit of all the target sequencing reads of each haplotype, generate the consensus sequence, compare the sequence of the tandem repeat unit of each target sequencing read with the consensus sequence, count the proportion of perfect matches, and obtain the sequence purity of the haplotype.

[0048] S62, calculate the mode, median, and mean of the copy number of all the target sequencing reads for each haplotype;

[0049] S63, select a preset number of the target sequencing reads whose copy number is near the mode, and display the sequence structure and variation information of the tandem repeat units in the target sequencing reads.

[0050] To achieve the above objective, according to a second aspect of the present invention, a target region tandem repeat variant typing device based on long-read sequencing is provided, the device comprising:

[0051] The alignment and screening module is set to acquire long-read sequencing data of the target region, align the long-read sequencing data to the reference genome, generate an alignment file, and screen out the sequencing reads located in the target region to obtain the target sequencing reads.

[0052] The parsing module is configured to parse the sequence of each of the target sequencing reads, identify the sequence of the tandem repeat unit in the target region, and obtain the position and copy number of the tandem repeat unit.

[0053] The detection module is configured to detect multi-source variation information within the target region based on the aforementioned comparison file, and obtain multi-source variation data.

[0054] The first module of haplotype typing is configured to use a haplotype construction algorithm, combined with the above copy number and the above multi-source variant data, to determine the haplotype attribution of each of the above target sequencing reads, and classify each of the above target sequencing reads into the first haplotype, the second haplotype, or the unassigned category according to the attribution, so as to obtain the preliminary haplotype typing results.

[0055] The second haplotype typing module is configured to redistribute the target sequencing reads in the unassigned categories based on the proportion and / or copy number distribution characteristics of the target sequencing reads in the first haplotype, the second haplotype, and the unassigned categories, to obtain the final haplotype typing result.

[0056] Among them, the above-mentioned multi-source variation information includes any of the following: single nucleotide variation, small fragment insertion / deletion and structural variation; the above-mentioned tandem repeat unit is one or more.

[0057] Furthermore, the above-mentioned parsing module includes:

[0058] The building unit is set as the regular expression or sequence pattern matching rule that constructs the above concatenated repeating unit;

[0059] The locating unit is configured to scan the target sequencing reads using the above regular expressions or sequence pattern matching rules to locate the positions of all matched tandem repeat units.

[0060] The copy number counting unit is set to count the number of tandem repeat units matched in each of the above target sequencing reads to obtain the above copy number.

[0061] Furthermore, when the long-read sequencing data is sequencing data of PCR products, the alignment and screening module further includes a verification unit, which is configured to verify the validity of the sequencing reads based on primer sequence matching degree after screening out the sequencing reads located in the target region and before obtaining the target sequencing reads, retain the sequencing reads that have passed the test, and obtain the target sequencing reads from the sequencing reads that have passed the test.

[0062] Furthermore, the aforementioned verification unit also includes a truncation module, configured to remove the primer sequences and flanking sequences at both ends of the aforementioned sequenced reads to obtain truncated reads, which are the aforementioned target sequenced reads.

[0063] Furthermore, the detection module includes: a first analysis unit configured to analyze the alignment file using a single nucleotide variant detection tool to extract single nucleotide variant site information and small fragment insertion / deletion site information within the target region; a second analysis unit configured to analyze the alignment file using a structural variant detection tool to extract structural variant site information within the target region; and an integration unit configured to integrate the single nucleotide variant site information, small fragment insertion / deletion site information, and structural variant site information to obtain the multi-source variant data.

[0064] Furthermore, the single nucleotide variant detection tools mentioned above are DeepVariant or longshot, and the structural variant detection tools mentioned above are Sawfish, cuteSV, or clair3.

[0065] Furthermore, the first haplotype typing module includes: a merging unit, configured to merge the copy number and the multi-source variant data to generate a comprehensive variant VCF file; and a haplotype phase inference unit, which inputs the comprehensive variant VCF file into the haplotype construction algorithm, performs haplotype phase inference on each of the target sequencing reads, and outputs the haplotype category to which each target sequencing read belongs.

[0066] Further, the second haplotype typing module includes: a first statistical unit, configured to count the number of target sequencing reads in the first haplotype, the second haplotype, and the unassigned category, and calculate the proportion of each to the total effective sequencing reads; a first allocation unit, configured to determine whether the proportion of the first haplotype and the proportion of the second haplotype are both greater than or equal to a first preset threshold; if so, the typing result is determined to be stable, no reallocation is performed, and the current typing result is taken as the final haplotype typing result; if not, the second allocation unit is executed; the second allocation unit is configured to determine whether the proportion of the first haplotype or the second haplotype is greater than or equal to a second preset threshold, and whether the proportion of the unassigned category is greater than or equal to a third preset threshold; if so, the unassigned category is recalibrated to the haplotype category with the smaller proportion among the first haplotype or the second haplotype, and the haplotype category with the smaller proportion is recalibrated to the unassigned category, and after obtaining a new typing result, the new haplotype category is assigned to the second allocation unit. The result is recorded as the final haplotype typing result; otherwise, the acquisition unit and the third allocation unit are executed sequentially. The acquisition unit is set to acquire the histogram of the copy number distribution of tandem repeat units in all the target sequencing reads, identify the copy number intervals corresponding to the two highest peaks, and use the midpoint of these two copy number intervals as the copy number threshold. The third allocation unit is set to classify all the target sequencing reads into two haplotypes according to the copy number threshold. The ratio of the two haplotypes is calculated, and it is determined whether the ratio of the two haplotypes is (4~10):1. If it is, the two haplotypes are the final haplotype typing result. If not, the fourth allocation unit is executed. The fourth allocation unit is set to randomly allocate the target sequencing reads into two haplotypes with a ratio of 1:1. The two haplotypes are the final haplotype typing result. The first preset threshold, the second preset threshold, and the third preset threshold are all 30%~35%.

[0067] Furthermore, the aforementioned apparatus also includes an output module, configured to generate a consensus sequence for each haplotype based on the tandem repeat unit sequence of the target sequencing reads contained therein, according to the final haplotype typing results, calculate the copy number statistics and sequence purity of each haplotype, and generate visualization results.

[0068] Furthermore, the above output module includes:

[0069] The alignment unit is configured to extract the sequence of the tandem repeat unit of all the target sequencing reads of each haplotype, generate the consensus sequence, align the sequence of the tandem repeat unit of each target sequencing read with the consensus sequence, count the proportion of perfect matches, and obtain the sequence purity of the haplotype.

[0070] The calculation unit is configured to calculate the mode, median, and mean of the copy number of all the target sequencing reads for each haplotype.

[0071] The display unit is set to filter out a preset number of the target sequencing reads whose copy number is near the mode, and to display the sequence structure and variation information of the tandem repeat units in the target sequencing reads.

[0072] To achieve the above objectives, according to a third aspect of the present invention, a computer-readable storage medium is provided, the storage medium including a stored program, wherein, when the program is executed, the device on which the storage medium is located executes the above-described target region tandem repeat variant genotyping method based on long read sequencing.

[0073] To achieve the above objectives, according to a fourth aspect of the present invention, a processor is provided for running a program, wherein the program executes the above-described target region tandem repeat variant genotyping method based on long-read sequencing.

[0074] By applying the technical solution of this invention, target sequencing reads are obtained by acquiring and comparing long-read sequencing data of the target region; sequence analysis is performed on each target sequencing read to identify the position and copy number of tandem repeat units; multi-source variant information (including at least two of single nucleotide variants, small fragment insertions / deletions, and structural variants) is detected based on the alignment file; using a haplotype construction algorithm, combined with copy number and multi-source variant data, haplotype classification is performed on each target sequencing read, classifying it into a first haplotype, a second haplotype, or an unassigned category; finally, based on the proportion and / or copy number distribution characteristics of reads in the two haplotypes and the unassigned category, the target sequencing reads in the unassigned category are reassigned to obtain the final haplotype typing result.

[0075] The above method overcomes the bottleneck of existing software in high-depth sequencing, which cannot perform single-molecule-level quantification and genotyping accuracy for known motifs, by detecting CPN and correcting haplotypes for each read. It significantly improves the genotyping accuracy of complex tandem repeat regions (such as long repeats or highly variable regions). In addition, this method can effectively distinguish between heterozygous and homozygous samples, and by calculating motif purity and visualizing sequence structure based on haplotypes, it provides high-resolution and highly consistent molecular genotyping basis for genetic disease research, forensic individual identification, and population genetic analysis. It solves the detection bias caused by the read length limitation of traditional second-generation sequencing and the problem of inaccurate read allocation in existing third-generation analysis tools targeting high-depth data. Attached Figure Description

[0076] The accompanying drawings, which form part of this specification, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0077] Figure 1 The CPN depth distribution of two haplotypes in the P3920 sample according to an embodiment of the present invention is shown; wherein, A is the haplotype with the larger mode and B is the haplotype with the smaller mode.

[0078] Figure 2 The CPN depth distribution of two haplotypes in the LY1242 sample according to an embodiment of the present invention is shown; wherein, A is the haplotype with the larger mode and B is the haplotype with the smaller mode.

[0079] Figure 3 The sequence structure L of 20 reads near the median of each haplotype for each sample according to an embodiment of the present invention is shown; where A is sample P3920 and B is sample LY1242.

[0080] Figure 4 A hardware block diagram of a target region tandem repeat variant typing based on long-read sequencing according to an embodiment of the present invention is shown.

[0081] Figure 5 A schematic diagram of a target region tandem repeat variant analysis device based on long-read sequencing according to an embodiment of the present invention is shown; wherein the content outlined in the dashed box represents the preferred technical solution. Detailed Implementation

[0082] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the embodiments.

[0083] Terminology Explanation:

[0084] Sequence purity: Used to assess the consistency of tandem repeat unit (motif) sequences within a haplotype. Specifically, for a given haplotype (e.g., Hap1), the system first generates a consensus sequence based on the motif sequences of all sequencing reads within that haplotype. Then, the system aligns the motif sequence of each read within that haplotype with the consensus sequence. If the motif sequence of a read perfectly matches the consensus sequence (i.e., there are no single nucleotide variants, small insertions, or deletions), the read is considered a "high-purity read." Sequence purity is the percentage or proportion of high-purity reads out of the total number of reads within that haplotype.

[0085] A motif (often translated as motif, motif, repeat unit, or core sequence) is the smallest and most repetitive basic DNA sequence unit in a tandem repeat sequence.

[0086] As mentioned in the background section, existing tandem repeat detection software based on third-generation sequencing is mostly designed for whole-genome or medium-depth sequencing data. When faced with high-depth targeted sequencing data, it is often unable to accurately analyze the copy number of each read for a known motif, and lacks a dynamic correction mechanism for haplotype attribution, resulting in significant deviations in the genotyping accuracy of complex tandem repeat regions and the allocation of unhappy reads.

[0087] In this invention, the applicant attempts to integrate multi-source variation information such as single nucleotide variants, small fragment insertions and deletions, and structural variations with tandem repeat copy numbers, and uses a haplotype construction algorithm to determine the haplotype attribution of each sequenced read. Furthermore, based on the haplotype ratio distribution and copy number distribution characteristics, the unassigned sequenced reads are reassigned, thereby achieving high-precision haplotype typing of tandem repeat variants in high-depth targeted data. Based on this, the protection scheme of this invention is proposed.

[0088] In a first typical embodiment of the present invention, a method for genotyping target region tandem repeat variants based on long-read sequencing is provided. This method includes: S1, acquiring long-read sequencing data of the target region, aligning the long-read sequencing data to a reference genome, generating an alignment file, and screening out sequencing reads located within the target region to obtain target sequencing reads; S2, performing sequence analysis on each target sequencing read to identify the sequence of the tandem repeat unit within the target region, obtaining the position and copy number of the tandem repeat unit; S3, based on the alignment file, detecting multi-source variant information within the target region to obtain multi-source variant data; S4, using a single... The ploidy construction algorithm, combining the copy number and multi-source variant data mentioned above, determines the haplotype assignment of each target sequencing read and classifies each target sequencing read into a first haplotype, a second haplotype, or an unassigned category based on the assignment, thus obtaining a preliminary haplotype typing result; S5, based on the proportional distribution and / or copy number distribution characteristics of the target sequencing reads in the first haplotype, second haplotype, and unassigned categories, the target sequencing reads in the unassigned categories are reassigned to obtain the final haplotype typing result; wherein, the multi-source variant information includes any combination of the following: single nucleotide variants, small fragment insertions / deletions, and structural variants; the tandem repeat unit is one or more.

[0089] The above method acquires target region data using long-read sequencing technology and uses alignment files to screen sequencing reads located within the target region, ensuring that subsequent analysis targets specific regions. Sequence parsing of each read identifies the position and copy number of tandem repeat units, achieving precise quantification at the single-molecule level. Simultaneously, based on the alignment files, multi-source variation information such as single nucleotide variants, small fragment insertions / deletions, and structural variations is detected, and combined with copy number data, a haplotype construction algorithm is used to determine the haplotype assignment of each read. For unassigned categories in the preliminary typing, reads are redistributed based on the proportional distribution and copy number distribution characteristics of the first haplotype, second haplotype, and unassigned categories, resulting in a more accurate final haplotype typing result. This workflow helps overcome the limitations of traditional short-read sequencing in typing complex tandem repeat regions and improves the accuracy of haplotype assignment in high-depth targeted sequencing data through haplotype phase inference combining multi-source variation information and copy number, effectively solving the problem of handling unassigned reads and providing high-resolution molecular typing evidence for genetic disease research and forensic identification.

[0090] In a preferred embodiment of the present invention, S2 includes the following steps: S21, constructing a regular expression or sequence pattern matching rule for the tandem repeat unit; S22, scanning the target sequencing reads using the regular expression or sequence pattern matching rule to locate the positions of all matched tandem repeat units; S23, counting the number of matched tandem repeat units in each target sequencing read to obtain the copy number.

[0091] By scanning target sequencing reads using regular expressions or sequence pattern matching rules, the start and end positions of each tandem repeat unit (motif) within a single read can be precisely located. The copy number (CPN) of each read in the target region is directly obtained by counting the number of matched units. This method is not only computationally efficient but also preserves the original sequence information of each read, providing a precise quantitative basis for subsequent haplotype typing and sequence purity calculations. It is suitable for target region analysis with known specific motif sequences.

[0092] In a preferred embodiment of the present invention, in S1 above, when the long read sequencing data is sequencing data of PCR products, after screening out the sequencing reads located in the target region and before obtaining the target sequencing reads, it is necessary to verify the validity of the sequencing reads based on primer sequence matching degree, retain the sequencing reads that have passed the sequencing test, and obtain the target sequencing reads from the sequencing reads that have passed the sequencing test.

[0093] The above steps, when the long-read sequencing data are PCR products, validate the sequencing reads based on primer sequence matching to ensure that the reads originate from the amplified products of the target region, thereby eliminating non-specific background noise. Retaining successfully sequenced reads helps improve the quality of subsequent data analysis, ensuring that the analyzed reads have a clear source from the target region and enhancing the reliability of the genotyping results.

[0094] In a preferred embodiment of the present invention, the step of obtaining the target sequencing reads from the above-mentioned sequencing reads includes: removing the primer sequences and flanking sequences at both ends of the above-mentioned sequencing reads to obtain truncated reads, which are the target sequencing reads.

[0095] The above steps, by removing the primer and flanking sequences at both ends of the sequencing reads, focus on the core sequence of the target tandem repeat region, avoiding interference from non-target sequences with the matching algorithm. The truncated reads are the target sequencing reads, further improving the accuracy and specificity of tandem repeat unit identification.

[0096] In a preferred embodiment of the present invention, S3 includes the following steps: S31, analyzing the alignment file using a single nucleotide variant detection tool to extract single nucleotide variant (SNP) information and small fragment insertion / deletion (Indel) site information in the target region; S32, analyzing the alignment file using a structural variant detection tool to extract structural variant (SV) site information in the target region; S33, integrating the single nucleotide variant site information, small fragment insertion / deletion site information, and structural variant site information to obtain the multi-source variant data.

[0097] The above steps utilize specialized variant detection tools to extract variant information at the single nucleotide, small fragment, and structural levels, respectively. Single nucleotide variant detection tools can identify point mutations in sequences, small fragment insertion / deletion detection tools can identify additions or subtractions in short sequences, and structural variant detection tools can identify large-scale sequence rearrangements or complex variations. Integrating these three types of variant information forms a rich multi-source variant dataset.

[0098] These variant sites are located on the flanks or inside the tandem repeat regions, forming unique genetic markers for each read. These markers play a crucial role in subsequent differentiation of different haplotypes because they provide the phase information required by haplotype construction algorithms, enabling the algorithms to distinguish reads from different chromosomal sources.

[0099] In a preferred embodiment of the present invention, the single nucleotide variant detection tool is DeepVariant or longshot, and the structural variant detection tool is Sawfish, cuteSV, or clair3.

[0100] The methods described above utilize validated and efficient variant detection tools. DeepVariant or longshot are suitable for high-precision detection of single nucleotide and small fragment variants, effectively reducing the impact of the high error rate in third-generation sequencing data on variant detection. Sawfish, cuteSV, or clair3 are suitable for structural variant detection in long-read data. This combination ensures the reliability and integrity of multi-source variant data, contributing to improved accuracy in subsequent haplotype typing.

[0101] In a preferred embodiment of the present invention, S4 includes the following steps: S41, merging the copy number and the multi-source variant data to generate a comprehensive variant VCF file; S42, inputting the comprehensive variant VCF file into the haplotype construction algorithm to perform haplotype phase inference on each target sequencing read and outputting the haplotype category to which each target sequencing read belongs.

[0102] The above steps integrate the copy number information of tandem repeat units with traditional SNP, Indel, and SV variant information to generate a comprehensive variant file containing rich genetic information. The haplotype construction algorithm utilizes the characteristic that long-read sequencing reads span multiple heterozygous sites, combined with the unique variant feature of copy number, to perform phase inference on each read.

[0103] Copy number information has a unique discriminative power in the classification of cascaded repeating regions because it directly reflects the difference in the number of repeating units. Incorporating copy number into the haplotype construction process helps the algorithm more accurately distinguish reads with similar sequence structures but different copy numbers, thereby improving the accuracy of haplotype attribution.

[0104] In a preferred embodiment of the present invention, S5 includes the following steps: S51, counting the number of target sequencing reads in the first haplotype, the second haplotype, and the unassigned category, and calculating the proportion of each to the total effective sequencing reads; S52, determining whether the proportion of the first haplotype and the proportion of the second haplotype are both greater than or equal to a first preset threshold; if so, determining that the genotyping result is stable, not performing reallocation, and taking the current genotyping result as the final haplotype genotyping result; if not, proceeding to S53;

[0105] S53, determine that the proportion of the first haplotype or the second haplotype is greater than or equal to a second preset threshold, and the proportion of the unassigned category is greater than or equal to a third preset threshold; if so, recalibrate the unassigned category to the haplotype category with the smaller proportion among the first haplotype and the second haplotype, and recalibrate the haplotype category with the smaller proportion to the unassigned category to obtain a new classification result, and record the new classification result as the final haplotype classification result; if not, execute S54;

[0106] S54, obtain the histogram of copy number distribution of tandem repeat units in all the above target sequencing reads, identify the copy number intervals corresponding to the first two highest peaks, and use the midpoint of these two copy number intervals as the copy number threshold.

[0107] S55, based on the above copy number threshold, classify all the above target sequencing reads into two haplotypes; calculate the ratio of the two haplotypes and determine whether the ratio of the two haplotypes is (4~10):1 (e.g., 4:1, 4.5:1, 5:1, 6:1, 7:1, 8:1, 9:1 or 10:1). If yes, the two haplotypes are the final haplotype typing results. If not, proceed to S56.

[0108] S56, Randomly assign the above target sequencing reads to two haplotypes in a 1:1 ratio, and the above two haplotypes are the final haplotype typing results;

[0109] The first, second, and third preset thresholds are all 30% to 35% (e.g., 30%, 31%, 32%, 33%, 34%, or 35%).

[0110] Step S5 described above provides an adaptive genotyping correction mechanism to address differences in genotyping results across different samples. Genotyping is considered stable when both haplotype ratios are high. When a large number of unassigned reads exist and the ratio of a particular haplotype is low, a permutation operation is used to rebalance the allocation and reduce noise interference. When the haplotype construction algorithm cannot effectively distinguish between haplotypes, the distribution characteristics of tandem repeat copy number are utilized to force genotyping through peak identification and threshold segmentation. If the genotyping ratio meets a specific criterion (e.g., 4:1), it is considered valid; otherwise, the sample is assumed to be homozygous, and reads are evenly distributed between the two haplotypes. This multi-level correction strategy helps improve genotyping robustness in high-depth, high-noise, or homozygous samples, ensuring the biological validity of the final genotyping results.

[0111] It should be noted that the peak height is not limited in S54 above. Furthermore, since a histogram is composed of multiple bins (bars), sometimes two peaks may be adjacent and indistinguishable. Therefore, to distinguish between the two true peaks, a peak spacing is set. In a preferred embodiment of the present invention, the peak spacing is set to 4~10 (e.g., 4, 5, 6, 7, 8, 9, or 10), more preferably 5 (referring to 5 bins).

[0112] The peak spacing is determined based on the pathogenicity of the repetitive sequences being studied. For example, under normal circumstances, the number of repetitions of a motif in a sample is 10. Some diseases reach the pathogenicity range with more than 50 repetitions, while others reach the pathogenicity range with more than 100 or even 200 repetitions. In these cases, the peak spacing can be set appropriately larger.

[0113] In a more preferred embodiment of the present invention, in S55, when the sequencing depth is high (e.g., 200× or higher), all the target sequencing reads are divided into two haplotypes according to the above copy number threshold; the ratio of the two haplotypes is calculated, and it is determined whether the ratio of the two haplotypes is (4~10):1 (e.g., 4:1, 4.5:1, 5:1, 6:1, 7:1, 8:1, 9:1 or 10:1). If yes, the two haplotypes are the final haplotype typing result. If no, S56 is executed.

[0114] It should be noted that the step S56 above, "randomly assigning the target sequencing reads to two haplotypes in a 1:1 ratio," specifically includes the following steps: First, the reads are randomly shuffled, and then a dividing point is set based on the total number of reads. If the total number of reads is even, all reads are randomly divided into two haplotypes. For example, if there are 100 reads in total (the total number of reads is even), the first 50 are allele1, and the rest are allele2.

[0115] If the total number of reads is odd, then the value of (total number of reads - 1) / 2 is used as the dividing point. Reads before (including) the dividing point are divided into allele1, and the rest are divided into allele2. For example, if there are a total of 101 reads (the total number of reads is odd), then the first 50 are still allele1, and the rest are allele2.

[0116] In a preferred embodiment of the present invention, after obtaining the final haplotype typing result, the target region tandem repeat variant typing method further includes: S6, based on the final haplotype typing result, for each haplotype, generating a consensus sequence based on the tandem repeat unit sequence of the target sequencing reads contained therein, calculating the copy number statistics and sequence purity of each haplotype, and generating a visualization result.

[0117] The above-mentioned S6 section provides in-depth analysis and visualization of the genotyping results. Sequence purity is calculated by generating consensus sequences, allowing assessment of the consistency of sequences within each haplotype. Copy number statistics provide the central tendency and dispersion of repeat length for that haplotype. Visualization visually presents the main repeat unit arrangement patterns and their potential subtle variations, aiding in understanding the complex structure of tandem repeat regions and their association with phenotypes or diseases.

[0118] In a preferred embodiment of the present invention, S6 includes the following steps: S61, extracting the sequences of the tandem repeat units of all the target sequencing reads for each haplotype, generating the consensus sequence, comparing the sequences of the tandem repeat units of each target sequencing read with the consensus sequence, calculating the proportion of perfect matches, and obtaining the sequence purity of the haplotype; S62, calculating the mode, median, and mean of the copy numbers of all the target sequencing reads for each haplotype; S63, selecting a predetermined number of target sequencing reads with copy numbers near the mode, and displaying the sequence structure and variation information of the tandem repeat units in the target sequencing reads.

[0119] The S6 step described above calculates sequence purity through the alignment unit to assess the consistency within haplotypes; provides a statistical description of copy number through the calculation unit; and offers visualization through the display unit. These functions collectively provide a comprehensive evaluation and intuitive presentation of the genotyping results, aiding in understanding the structural characteristics and distribution of tandem repeat variants.

[0120] In a second typical embodiment of the present invention, a target region tandem repeat variant typing device based on long-read sequencing is provided (see...). Figure 5The device includes: an alignment and screening module 10, configured to acquire long-read sequencing data of a target region, align the long-read sequencing data to a reference genome, generate an alignment file, and screen out sequencing reads located within the target region to obtain target sequencing reads; a parsing module 20, configured to perform sequence parsing on each target sequencing read, identify the sequence of tandem repeat units within the target region, and obtain the position and copy number of the tandem repeat units; a detection module 30, configured to detect multi-source variant information within the target region based on the alignment file to obtain multi-source variant data; and a haplotype typing first module 40, configured to use a haplotype construction algorithm, combined with the copy number... Based on the aforementioned multi-source variant data, haplotype attribution is determined for each target sequencing read, and each target sequencing read is classified into a first haplotype, a second haplotype, or an unassigned category according to its attribution, thus obtaining preliminary haplotype typing results. The second haplotype typing module 50 is configured to redistribute the target sequencing reads in the unassigned categories based on the proportional distribution and / or copy number distribution characteristics of the target sequencing reads in the first haplotype, the second haplotype, and the unassigned categories, thus obtaining final haplotype typing results. The aforementioned multi-source variant information includes any combination of the following: single nucleotide variants, small fragment insertions / deletions, and structural variants; the aforementioned tandem repeat units are one or more types.

[0121] The aforementioned device implements the modular functionality of the method described above. The comparison and screening module 10 is responsible for data preprocessing and validity filtering; the parsing module 20 is responsible for the quantification of tandem repeat units; the detection module 30 is responsible for the extraction of multi-source variants; the first haplotype typing module 40 is responsible for preliminary haplotype phase inference; and the second haplotype typing module 50 is responsible for typing correction based on statistical characteristics. This device encapsulates the complex bioinformatics analysis process into independent modules, which helps improve the standardization and reproducibility of the analysis process and is suitable for high-throughput tandem repeat variant typing of clinical or research samples.

[0122] In a preferred embodiment of the present invention, the parsing module 20 includes: a construction unit 201, configured to construct a regular expression or sequence pattern matching rule for the tandem repeat unit; a positioning unit 202, configured to scan the target sequencing reads using the regular expression or sequence pattern matching rule to locate the positions of all matched tandem repeat units; and a copy number counting unit 203, configured to count the number of matched tandem repeat units in each target sequencing read to obtain the copy number.

[0123] The aforementioned parsing module 20 defines matching rules through the construction unit 201, accurately identifies the positions of tandem repeating units through the positioning unit 202, and finally calculates the copy number using the copy number counting unit 203. This modular design makes the parsing process flexible and efficient, adaptable to different motif sequences in different target regions, and ensures the accuracy of copy number calculation.

[0124] In a preferred embodiment of the present invention, when the long read sequencing data is sequencing data of PCR products, the alignment and screening module 10 further includes a verification unit 101, which is configured to perform validity verification on the sequencing reads based on primer sequence matching degree after screening out the sequencing reads located in the target region and before obtaining the target sequencing reads, retain the sequencing reads that have passed the test, and obtain the target sequencing reads from the sequencing reads that have passed the test.

[0125] When the data is PCR product, the above-mentioned verification unit 101 verifies the validity of sequencing reads based on primer sequence matching degree to ensure that the sequencing reads originate from the target region amplification product, thereby eliminating non-specific background noise and improving the quality of the analysis data.

[0126] In a preferred embodiment of the present invention, the above-mentioned verification unit 101 further includes: a truncation module 1011, which is configured to remove the primer sequences and flanking sequences at both ends of the above-mentioned sequencing reads that have passed the test, to obtain truncated reads, which are the above-mentioned target sequencing reads.

[0127] The aforementioned extraction module 1011 removes the primer sequences and flanking sequences at both ends of the sequencing reads, focusing on the core sequence of the target tandem repeat region. This avoids interference from non-target sequences with the matching algorithm and improves the accuracy of tandem repeat unit identification.

[0128] In a preferred embodiment of the present invention, the detection module 30 includes: a first analysis unit 301, configured to analyze the alignment file using a single nucleotide variant detection tool to extract single nucleotide variant site information and small fragment insertion / deletion site information in the target region; a second analysis unit 302, configured to analyze the alignment file using a structural variant detection tool to extract structural variant site information in the target region; and an integration unit 303, configured to integrate the single nucleotide variant site information, small fragment insertion / deletion site information, and structural variant site information to obtain the multi-source variant data.

[0129] The aforementioned detection module 30 processes different types of variants through the first analysis unit 301 and the second analysis unit 302, respectively, and finally the results are merged by the integration unit 303. This clearly defined division of labor design can fully utilize the advantages of specialized tools, improve the comprehensiveness and accuracy of variant detection, and provide rich characteristic data for haplotype typing.

[0130] In a preferred embodiment of the present invention, the single nucleotide variant detection tool is DeepVariant or Longshot, and the structural variant detection tool is Sawfish, CuteSV, or Clair3. The device utilizes validated and efficient variant detection tools; DeepVariant or Longshot is suitable for high-precision detection of single nucleotide and small fragment variants, while Sawfish, CuteSV, or Clair3 is suitable for structural variant detection in long-read data. This combination ensures the reliability and integrity of multi-source variant data, contributing to improved accuracy in subsequent haplotype typing.

[0131] In a preferred embodiment of the present invention, the first haplotype typing module 40 includes: a merging unit 401, configured to merge the copy number and the multi-source variant data to generate a comprehensive variant VCF file; and a haplotype phase inference unit 402, which inputs the comprehensive variant VCF file into the haplotype construction algorithm, performs haplotype phase inference on each of the target sequencing reads, and outputs the haplotype category to which each target sequencing read belongs.

[0132] The first haplotype typing module 40 integrates copy number and variation information through the merging unit 401 to generate a comprehensive input file, and the haplotype phase inference unit 402 performs phase inference using the haplotype construction algorithm. This design introduces copy number information into the haplotype construction process, enhances the ability to distinguish between tandem repeat regions, and helps to more accurately classify reads into the correct haplotype.

[0133] In a preferred embodiment of the present invention, the haplotype typing second module 50 includes: a first statistics unit 501, configured to count the number of target sequencing reads in the first haplotype, the second haplotype, and the unassigned category, and calculate the proportion of each to the total effective sequencing reads; a first allocation unit 502, configured to determine whether the proportion of the first haplotype and the proportion of the second haplotype are both greater than or equal to a first preset threshold; if so, the typing result is determined to be stable, no reallocation is performed, and the current typing result is taken as the final haplotype typing result; if not, the second allocation unit 503 is executed.

[0134] The second allocation unit 503 is configured to determine whether the proportion of the first haplotype or the second haplotype is greater than or equal to a second preset threshold, and whether the proportion of the unallocated category is greater than or equal to a third preset threshold. If so, the unallocated category is recalibrated to the haplotype category with the smaller proportion among the first haplotype and the second haplotype, and the haplotype category with the smaller proportion is recalibrated to the unallocated category. After obtaining a new classification result, the new classification result is recorded as the final haplotype classification result. If not, the acquisition unit 504 and the third allocation unit 505 are executed sequentially.

[0135] The aforementioned acquisition unit 504 is configured to acquire the histogram of the copy number distribution of tandem repeat units in all the aforementioned target sequencing reads, identify the copy number intervals corresponding to the two highest peaks, and use the midpoint of these two copy number intervals as the copy number threshold.

[0136] The third allocation unit 505 is configured to classify all the target sequencing reads into two haplotypes based on the copy number threshold; calculate the ratio of the two haplotypes and determine whether the ratio of the two haplotypes is (4~10):1 (e.g., 4:1, 4.5:1, 5:1, 6:1, 7:1, 8:1, 9:1 or 10:1). If yes, the two haplotypes are the final haplotype typing result; otherwise, the fourth allocation unit 506 is executed.

[0137] The fourth allocation unit 506 is configured to randomly allocate the target sequencing reads into two haplotypes in a 1:1 ratio, and the two haplotypes are the final haplotype result; wherein the first preset threshold, the second preset threshold and the third preset threshold are all 30%~35% (e.g., 30%, 31%, 32%, 33%, 34% or 35%).

[0138] The aforementioned second haplotype typing module 50 calculates the proportions using the first statistical unit 501, and then coordinates with the first allocation unit 502, the second allocation unit 503, the acquisition unit 504, the third allocation unit 505, and the fourth allocation unit 506 to execute multi-level typing correction logic. This complex logic design aims to handle various typing anomalies, including proportion imbalances, excessive unallocated reads, and overlapping copy number distributions. Through adaptive threshold judgment and redistribution strategies, it ensures the accuracy and biological significance of the final typing results.

[0139] In a preferred embodiment of the present invention, the above-mentioned apparatus further includes: an output module 60, configured to generate a consensus sequence for each haplotype based on the tandem repeat unit sequence of the target sequencing reads contained therein, according to the final haplotype typing result, calculate the copy number statistics and sequence purity of each haplotype, and generate a visualization result.

[0140] The output module 60 performs in-depth analysis and visualization of the typing results, calculates sequence purity by generating consensus sequences, and evaluates the consistency within haplotypes; provides a statistical description of copy number through the calculation unit 602; and intuitively presents the arrangement pattern of repeating units through visualization, providing researchers with a comprehensive evaluation basis.

[0141] In a preferred embodiment of the present invention, the output module 60 includes: an alignment unit 601, configured to extract the sequences of the tandem repeat units of all target sequencing reads for each haplotype, generate the consensus sequence, align the sequences of the tandem repeat units of each target sequencing read with the consensus sequence, calculate the proportion of perfect matches, and obtain the sequence purity of the haplotype; a calculation unit 602, configured to calculate the mode, median, and mean of the copy numbers of all target sequencing reads for each haplotype; and a display unit 603, configured to select a preset number of target sequencing reads with copy numbers near the mode, and display the sequence structure and variation information of the tandem repeat units in the target sequencing reads.

[0142] The output module 60 calculates sequence purity and assesses the consistency within haplotypes through the alignment unit 601; provides a statistical description of copy number through the calculation unit 602; and performs visualization through the display unit 603. These functions together provide a comprehensive evaluation and intuitive display of the genotyping results, which helps to understand the structural characteristics and distribution of tandem repeat variants.

[0143] In a third typical embodiment of the present invention, a computer-readable storage medium is provided, the storage medium including a stored program, wherein, when the program is executed, the device on which the storage medium is located executes the above-described target region tandem repeat variant typing method based on long read sequencing.

[0144] The aforementioned computer-readable storage medium stores program code for executing the method of this application. When the program runs on a computer device, it can automatically execute the entire process from data preprocessing, variant detection, haplotype typing to result analysis. This implementation allows the method to be integrated into a bioinformatics analysis platform, improving analysis efficiency, reducing manual intervention, ensuring the reproducibility and standardization of the analysis process, and making it suitable for large-scale genomics research or clinical diagnostic applications.

[0145] In a fourth typical embodiment of the present invention, a processor is provided for running a program, wherein the program executes the above-described target region tandem repeat variant typing method based on long read sequencing.

[0146] The method provided in this application can be executed on a terminal, computer terminal, or similar computing device. Taking running on a terminal as an example, Figure 4 This is a hardware structure block diagram of a target region tandem repeat variant typing based on long read sequencing according to an embodiment of the present invention.

[0147] like Figure 4 As shown, a terminal may include one or more ( Figure 4 Only one is shown in the diagram. A processor A1 (processor A1 may include, but is not limited to, a microprocessor (MCU) or a programmable logic device (FPGA), etc.) and a memory B1 for storing data are also shown. Optionally, the terminal may further include a transmission device C1 for communication functions and an input / output device D1. Those skilled in the art will understand that... Figure 4 The structure shown is for illustrative purposes only and does not limit the structure of the terminal described above. For example, the terminal may also include components that are larger than... Figure 4 The more or fewer components shown, or having the same Figure 4 The different configurations shown.

[0148] Memory B1 can be used to store computer programs, such as application software programs and modules, like the computer programs corresponding to methods such as comparison, sliding window, calculation, and lookup tagging in the embodiments of the present invention. Processor A1 executes various functional applications and data processing by running the computer programs stored in memory B1, thereby implementing the methods described above. Memory B1 may include high-speed random access memory and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, memory B1 may further include memory remotely located relative to processor A1, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0149] Transmission device C1 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the terminal's communication provider. In one example, transmission device C1 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, transmission device C1 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0150] Obviously, those skilled in the art should understand that some modules or steps of this application described above can be implemented in general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device, or fabricating them separately as individual integrated circuit modules, or fabricating multiple modules or steps into a single integrated circuit module. Thus, this application is not limited to any particular hardware and software combination.

[0151] The present invention will be further described in detail below with reference to specific embodiments, which should not be construed as limiting the scope of protection claimed by the present invention.

[0152] Example 1

[0153] This invention relates to sequencing and genotyping of PCR product data from multiple normal and disease samples. Taking the normal sample LY1242 and the disease sample P3920 as examples, the specific method is as follows:

[0154] 1. Based on the given primer and barcode sequences, split the PCR product third-generation sequencing data to obtain the sequencing data for each sample.

[0155] 2. Obtain the alignment BAM file from the reference genome of hg38 (Human Genome Assembly GRCh38, which is currently the gold standard reference sequence for human genome research) that the sequencing data is aligned with.

[0156] 3. Use a self-written program to analyze and compare the BAM file, and perform specific detection of one or more motifs on the reads of the target interval.

[0157] First, the beginning and end sequences of each read are compared with the known primer sequences to ensure they are PCR-compatible reads. The primers and flanking sequences at both ends of the compatible reads are truncated. A regular expression matching algorithm is used to obtain the specific location of each motif (in this example, two motifs are involved: CTA and CTG) on each truncated read. The output includes the read ID, the truncated sequence, each position of the motif, and the number of motifs (i.e., copy number, CPN value). It should be noted that when there are multiple motifs, the CPN value for each motif is calculated separately. Later, during genotyping, either the CPN of a primary motif or the sum of several specific motifs can be used. Generally, the analysis focuses on motifs specifically related to the spread of a particular disease. For example, in this case, the main motif is CTG, which, when it repeatedly spreads to a certain threshold, becomes pathogenic.

[0158] 4. Obtain variation information at the single-molecule level and perform read typing.

[0159] SNP / Indel detection: Single nucleotide variants (SNPs) and small fragment insertions / deletions (Indels) are detected using DeepVariant-aligned BAM files (also known as BAM files), generating corresponding VCF files.

[0160] Structural variation (SV) detection: Use Sawfish to perform structural variation (SV) detection on the same BAM file to obtain an SVCF file.

[0161] Tandem Repeat (TR) Detection: Use TRGT to perform tandem repeat (TR) detection on the same BAM file and generate a TR VCF file.

[0162] Haplotype / read typing: Based on the VCF results of the above three types of variant detection, HiPhase is used to perform haplotype typing on each sequencing read to obtain the haplotype assignment information of each read.

[0163] Segmentation efficiency evaluation: Based on the HiPhase output, the proportion of separable reads, including the proportions of hap1, hap2, and unhap, is statistically analyzed to evaluate the overall segmentation effect.

[0164] 5. Use a self-written program for classification correction.

[0165] If the classification ratio of both haps is ≥30% in the results of the above steps, no correction is performed. It should be noted that the "two haps" refer to the ratio of hap1 (allele1) and hap2 (allele2), while unhap is the unclassified read. Theoretically, most reads should be classified, so the proportion of unhap should be minimized. However, the purpose of this application is to prioritize ensuring that hap1 and hap2 have sufficient reads; therefore, when both hap1 and hap2 meet the requirement of ≥30%, the standard is considered met, and no correction is performed.

[0166] Otherwise, if the proportion of both hap (hap1 or hap2) and unhap is ≥30%, then unhap is recalibrated to the hap with the smaller proportion of the two haps, and the smaller hap is recalibrated to unhap. For example, if the proportion of hap1 is less than 30%, while the proportion of unhap is ≥30%, then hap1 is renamed unhap, and unhap is renamed hap1.

[0167] Otherwise, based on the CPN value distribution of one or more specific motifs, the highest peak of the first two distributions is obtained. The midpoint of the two peaks is used as the threshold to distinguish between the two haplotypes. The ratio of the two haplotypes is calculated. If the set standard (default 4:1) is met, the typing result is output as the final typing result. Otherwise, the sample is considered to be a complete homozygosity, and the reads are randomly assigned to two haplotypes with a ratio close to 1:1 as the final typing result.

[0168] 6. Analysis based on haplotypes.

[0169] Calculate the CPN mode, median, and mean of all reads for a specific motif (e.g., CTA or CTG) for each haplotype; calculate the purity of the motif; and visualize the sequence structure by selecting n reads (default 20 reads) of each haplotype near a specific mode. See [link to documentation] for details. Figure 1 , Figure 2 and Figure 3 Tables 1, 2, and 3 present data on haplotypes that have expanded (typically hap2). Since CTG can expand and cause pathological changes, Tables 1, 2, and 3 focus on CTG.

[0170] Table 1: Subtype proportions for each sample

[0171]

[0172] Table 2: Depth and other related statistics for each sample

[0173]

[0174] Table 3: Median purity and mean purity of each sample motif

[0175]

[0176] As can be seen from the above description, the embodiments of the present invention achieve the following technical effects:

[0177] This invention achieves high-precision analysis of tandem repeat (TR) sequences in target regions by combining the advantages of long reads from third-generation sequencing with a multi-tool synergistic single-molecule variant detection strategy.

[0178] This application achieves precise genotyping at the single-molecule level. Unlike traditional methods that only output overall genotyping results at the sample level, this method can trace and distinguish the haplotype origin of each sequencing read, effectively solving the problem of genotyping confusion or misassignment caused by similar lengths or sequences of two haplotypes at extremely high sequencing depths, and significantly improving the resolution and accuracy of genotyping.

[0179] This application overcomes the technical limitations of short-read sequencing. Addressing the detection bias caused by the limited read lengths of traditional next-generation sequencing (NGS) which makes it difficult to traverse complex or long tandem repeat regions, this invention utilizes third-generation sequencing data to directly cover the complete TR region. Furthermore, by comprehensively utilizing multi-source variation information (e.g., SNP / Indel / SV / TR), the robustness of haplotype construction is enhanced, thereby more reliably resolving the genetic variation structure within complex genomic regions.

[0180] This application improves the efficiency and completeness of targeted high-depth data analysis. The method is specifically designed for high-depth targeted sequencing data enriched by PCR, enabling efficient processing of massive amounts of read data while ensuring quality control of individual read sequences (such as primer sequence verification). This provides a means of analysis with both high depth coverage and high haplotype resolution for genetic research, disease mechanism analysis, and forensic identification.

[0181] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for genotyping tandem repeat variants in target regions based on long-read sequencing, characterized in that, The target region tandem repeat variant typing method includes: S1, acquire long-read sequencing data of the target region, align the long-read sequencing data to the reference genome, generate an alignment file, and screen out sequencing reads located in the target region to obtain target sequencing reads; S2, perform sequence parsing on each target sequencing read, identify the sequence of tandem repeat units in the target region, and obtain the position and copy number of the tandem repeat units; S3, based on the comparison file, detect multi-source variation information within the target region to obtain multi-source variation data; S4. Using the haplotype construction algorithm, combined with the copy number and the multi-source variant data, the haplotype attribution of each target sequencing read is determined, and each target sequencing read is classified into the first haplotype, the second haplotype, or the unassigned category according to the attribution, to obtain the preliminary haplotype typing results. S5, based on the proportion and / or copy number distribution characteristics of the target sequencing reads in the first haplotype, the second haplotype, and the unassigned category, the target sequencing reads in the unassigned category are redistributed to obtain the final haplotype typing result; The multi-source variation information includes any combination of the following: single nucleotide variants, small fragment insertions / deletions, and structural variations; The tandem repeating unit can be one or more.

2. The target region tandem repeat mutation typing method according to claim 1, characterized in that, S2 includes the following steps: S21, Construct the regular expression or sequence pattern matching rule for the serial repeating unit; S22, the target sequencing reads are scanned using the regular expression or sequence pattern matching rules to locate the positions of all matched tandem repeat units; S23, count the number of tandem repeat units matched in each of the target sequencing reads to obtain the copy number.

3. The target region tandem repeat variant typing method according to claim 2, characterized in that, In step S1, when the long-read sequencing data is sequencing data of PCR products, after screening out the sequencing reads located in the target region and before obtaining the target sequencing reads, it is necessary to verify the validity of the sequencing reads based on primer sequence matching degree, retain the sequencing reads that have passed the sequencing test, and obtain the target sequencing reads from the sequencing reads that have passed the sequencing test.

4. The target region tandem repeat variant typing method according to claim 3, characterized in that, The steps for obtaining the target sequencing reads from the sequenced reads include: removing the primer sequences and flanking sequences from both ends of the sequenced reads to obtain truncated reads, which are the target sequencing reads.

5. The target region tandem repeat variant typing method according to any one of claims 1-4, characterized in that, S3 includes the following steps: S31, use a single nucleotide variant detection tool to analyze the alignment file and extract single nucleotide variant site information and small fragment insertion / deletion site information within the target region; S32, use a structural variation detection tool to analyze the comparison file and extract structural variation site information within the target region; S33, integrate the single nucleotide variant site information, small fragment insertion / deletion site information, and structural variant site information to obtain the multi-source variant data.

6. The target region tandem repeat mutation typing method according to claim 5, characterized in that, The single nucleotide variant detection tool is DeepVariant or longshot, and the structural variant detection tool is Sawfish, cuteSV, or clair3.

7. The target region tandem repeat variant typing method according to any one of claims 1-4, characterized in that, S4 includes the following steps: S41, merge the copy number and the multi-source mutation data to generate a comprehensive mutation VCF file; S42, input the integrated variant VCF file into the haplotype construction algorithm, perform haplotype phase inference on each target sequencing read, and output the haplotype category to which each target sequencing read belongs.

8. The target region tandem repeat variant typing method according to any one of claims 1-4, characterized in that, S5 includes the following steps: S51, count the number of target sequencing reads in the first haplotype, the second haplotype and the unassigned category, and calculate the proportion of each to the total effective sequencing reads; S52, determine whether the proportions of the first haplotype and the second haplotype are both greater than or equal to a first preset threshold; if yes, determine that the typing result is stable, do not redistribute, and take the current typing result as the final haplotype typing result; if no, execute S53; S53, determine that the proportion of the first haplotype or the second haplotype is greater than or equal to a second preset threshold, and the proportion of the unassigned category is greater than or equal to a third preset threshold; if yes, then recalibrate the unassigned category to the haplotype category with the smaller proportion among the first haplotype and the second haplotype, and recalibrate the haplotype category with the smaller proportion to the unassigned category to obtain a new typing result, and record the new typing result as the final haplotype typing result; if no, execute S54; S54, obtain the histogram of copy number distribution of tandem repeat units in all the target sequencing reads, identify the copy number intervals corresponding to the first two highest peaks, and use the midpoint of these two copy number intervals as the copy number threshold. S55, based on the copy number threshold, classify all the target sequencing reads into two haplotypes; calculate the ratio of the two haplotypes, and determine whether the ratio of the two haplotypes is (4~10):

1. If yes, the two haplotypes are the final haplotype typing result; otherwise, proceed to S56. S56, The target sequencing reads are randomly assigned to two haplotypes in a 1:1 ratio, and the two haplotypes are the final haplotype typing results; The first preset threshold, the second preset threshold, and the third preset threshold are all 30% to 35%.

9. The target region tandem repeat variant typing method according to any one of claims 1-4, characterized in that, After obtaining the final haplotype typing result, the target region tandem repeat variant typing method further includes: S6. Based on the final haplotype typing results, for each haplotype, a consensus sequence is generated based on the tandem repeat unit sequences of the target sequencing reads contained therein. The copy number statistics and sequence purity of each haplotype are calculated, and a visualization result is generated.

10. The target region tandem repeat mutation typing method according to claim 9, characterized in that, S6 includes the following steps: S61, extract the sequence of the tandem repeat unit of all the target sequencing reads for each haplotype, generate the consensus sequence, compare the sequence of the tandem repeat unit of each target sequencing read with the consensus sequence, count the proportion of perfect matches, and obtain the sequence purity of the haplotype. S62, calculate the mode, median, and mean of the copy number for all the target sequencing reads for each haplotype; S63, filter out a preset number of target sequencing reads whose copy number is near the mode, and display the sequence structure and variation information of the tandem repeat units in the target sequencing reads.

11. A target region tandem repeat variant typing device based on long-read sequencing, characterized in that, The device includes: The alignment and screening module is configured to acquire long-read sequencing data of the target region, align the long-read sequencing data to the reference genome, generate an alignment file, and screen out sequencing reads located in the target region to obtain the target sequencing reads. The parsing module is configured to parse the sequence of each target sequencing read, identify the sequence of tandem repeat units within the target region, and obtain the position and copy number of the tandem repeat units; The detection module is configured to detect multi-source variation information within the target region based on the comparison file, and obtain multi-source variation data; The first haplotype typing module is configured to use a haplotype construction algorithm, combined with the copy number and the multi-source variant data, to determine the haplotype attribution of each target sequencing read, and classify each target sequencing read into the first haplotype, the second haplotype, or the unassigned category according to the attribution, so as to obtain the preliminary haplotype typing results. The second haplotype typing module is configured to redistribute the target sequencing reads in the unassigned categories based on the proportion and / or copy number distribution characteristics of the target sequencing reads in the first haplotype, the second haplotype, and the unassigned categories, to obtain the final haplotype typing result; The multi-source variation information includes any combination of the following: single nucleotide variants, small fragment insertions / deletions, and structural variations; The tandem repeating unit can be one or more.

12. The target region tandem repeating variation typing device according to claim 11, characterized in that, The parsing module includes: The building unit is configured as a regular expression or sequence pattern matching rule for constructing the concatenated repeating unit; The locating unit is configured to scan the target sequencing reads using the regular expression or sequence pattern matching rules to locate the positions of all matched tandem repeat units; The copy number counting unit is configured to count the number of tandem repeat units matched in each of the target sequencing reads to obtain the copy number.

13. The target region tandem repeating variation typing device according to claim 12, characterized in that, When the long-read sequencing data is sequencing data of PCR products, the alignment and screening module further includes: The verification unit is configured to perform validity verification on the sequencing reads based on primer sequence matching degree after screening out the sequencing reads located in the target region and before obtaining the target sequencing reads, retain the sequencing reads that have passed the test, and obtain the target sequencing reads from the sequencing reads that have passed the test.

14. The target region tandem repeating variation typing device according to claim 13, characterized in that, The verification unit further includes: The truncation module is configured to remove the primer sequences and flanking sequences at both ends of the sequencing reads to obtain truncated reads, which are the target sequencing reads.

15. The target region tandem repeating variation typing device according to claim 11, characterized in that, The detection module includes: The first analysis unit is configured to analyze the alignment file using a single nucleotide variant detection tool to extract single nucleotide variant site information and small fragment insertion / deletion site information within the target region; The second analysis unit is configured to use a structural variation detection tool to analyze the alignment file and extract structural variation site information within the target region; The integration unit is configured to integrate the single nucleotide variant site information, small fragment insertion / deletion site information, and the structural variant site information to obtain the multi-source variant data.

16. The target region tandem repeating variation typing device according to claim 15, characterized in that, The single nucleotide variant detection tool is DeepVariant or longshot, and the structural variant detection tool is Sawfish, cuteSV, or clair3.

17. The target region tandem repeating variation typing device according to claim 11, characterized in that, The first module of haplotype typing includes: The merging unit is configured to merge the copy number and the multi-source mutation data to generate a composite mutation VCF file; The haplotype phase inference unit inputs the integrated variant VCF file into the haplotype construction algorithm, performs haplotype phase inference on each target sequencing read, and outputs the haplotype category to which each target sequencing read belongs.

18. The target region tandem repeating variation typing device according to claim 11, characterized in that, The second module of haplotype typing includes: The first statistical unit is configured to count the number of target sequencing reads in the first haplotype, the second haplotype, and the unassigned category, and calculate the proportion of each to the total effective sequencing reads. The first allocation unit is configured to determine whether the proportion of the first haplotype and the proportion of the second haplotype are both greater than or equal to a first preset threshold; if so, the classification result is determined to be stable, no reallocation is performed, and the current classification result is taken as the final haplotype classification result; if not, the second allocation unit is executed. The second allocation unit is configured to determine whether the proportion of the first haplotype or the second haplotype is greater than or equal to a second preset threshold, and whether the proportion of the unallocated category is greater than or equal to a third preset threshold. If so, the unallocated category is recalibrated to the haplotype category with the smaller proportion among the first haplotype and the second haplotype, and the haplotype category with the smaller proportion is recalibrated to the unallocated category. After obtaining a new classification result, the new classification result is recorded as the final haplotype classification result. If not, the acquisition unit and the third allocation unit are executed sequentially. The acquisition unit is configured to acquire the histogram of the copy number distribution of tandem repeat units in all the target sequencing reads, identify the copy number intervals corresponding to the two highest peaks, and use the midpoint of these two copy number intervals as the copy number threshold. The third allocation unit is configured to classify all target sequencing reads into two haplotypes based on the copy number threshold; calculate the ratio of the two haplotypes; determine whether the ratio of the two haplotypes is (4~10):1; if so, the two haplotypes are the final haplotype typing result; otherwise, execute the fourth allocation unit. The fourth allocation unit is configured to randomly allocate the target sequencing reads into two haplotypes in a 1:1 ratio, and the two haplotypes are the final haplotype result. The first preset threshold, the second preset threshold, and the third preset threshold are all 30% to 35%.

19. The target region tandem repeating variation typing device according to claim 11, characterized in that, The device further includes: The output module is configured to generate a consensus sequence for each haplotype based on the tandem repeat unit sequences of the target sequencing reads contained therein, according to the final haplotype typing results, calculate the copy number statistics and sequence purity of each haplotype, and generate visualization results.

20. The target region tandem repeating variation typing device according to claim 19, characterized in that, The output module includes: The alignment unit is configured to extract the sequence of the tandem repeat unit of all the target sequencing reads for each haplotype, generate the consensus sequence, align the sequence of the tandem repeat unit of each target sequencing read with the consensus sequence, count the proportion of perfect matches, and obtain the sequence purity of the haplotype. The calculation unit is configured to calculate the mode, median, and mean of the copy number of all the target sequencing reads for each haplotype; The display unit is configured to filter out a preset number of target sequencing reads with copy numbers near the mode, and display the sequence structure and variation information of tandem repeat units in the target sequencing reads.

21. A computer-readable storage medium, characterized in that, The storage medium includes a stored program, wherein, when the program is executed, the device containing the storage medium is controlled to perform the target region tandem repeat variant typing method based on long read sequencing as described in any one of claims 1 to 10.

22. A processor, characterized in that, The processor is used to run a program, wherein the program executes the target region tandem repeat variant typing method based on long read sequencing as described in any one of claims 1 to 10.

Images