Gene haplotype typing method and device based on sequencing data and medium

Abstract

The invention belongs to the technical field of allele typing, and discloses a gene haplotype typing method and device based on sequencing data and a medium, and the method comprises the following steps: obtaining to-be-detected gene sequencing data from a to-be-detected sample; splitting the sequencing data into genes according to a reference sequence, and performing variation information detection and variation information filtering; after sequence splitting is carried out on the basis of the sequence obtained after variation information filtering, the haplotype sequence obtained after sequence splitting is corrected, and a to-be-detected gene haplotype typing result of the to-be-detected sample is obtained. According to the method, complex calculation is simplified by adopting a mode of firstly splitting sequences into genes, the method does not completely depend on sequence comparison results, and similar typing is finally evaluated by adopting a mode of combining comparison with backtracking variation information scoring, so that typing difference information can be effectively refined, and interference of sequencing data noise on the results is reduced; and a haplotype result is verified for multiple times by adopting a primary typing and correction mode, so that the result accuracy is higher.

Description

Technical Field

[0001] The present invention belongs to the technical field of allele typing, and in particular, relates to a gene haplotype typing method, device and medium based on sequencing data. Background Art

[0002] Human leukocyte antigen (HLA) is a polymorphic region of about 3.6M on the short arm of chromosome 6 that participates in immune response. In stem cell and organ transplantation, the consistency of the haplotype of the recipient and donor HLA gene is related to the success rate of transplantation and the severity of rejection. It has been reported that HLA is associated with more than 100 diseases, mainly autoimmune diseases, such as diabetes, rheumatoid arthritis, asthma, etc. Therefore, obtaining more and more accurate HLA haplotypes is of great significance for transplant matching and disease research. In recent years, the application of second-generation sequencing technology, such as HLA gene targeted region capture sequencing technology and whole exome sequencing technology (WES) for HLA gene typing has become relatively mature. The second-generation sequencing technology can obtain higher resolution than the traditional PCR method, but due to the shorter sequencing fragments and the relatively poor coverage and accuracy in high GC and other regions, it is more difficult in subsequent bioinformatics analysis, so it may not be possible to type all genes during typing, and there may be certain errors. Nanopore sequencing has the characteristic of obtaining long-fragment sequencing sequences, so the analysis and processing is relatively simple during bioinformatics analysis, and it is easier to obtain comprehensive and accurate haplotype information. In addition, it has the characteristics of simple sample processing, fast sequencing speed and low capital investment that other non-nanopore single-molecule long-fragment sequencing technologies do not have. In theory, it is more suitable for fast and accurate HLA typing. However, due to the relatively short commercialization time of nanopore sequencing technology, there are few existing HLA genotyping tools based on nanopore sequencing technology, or the algorithm cannot be well developed based on the characteristics of nanopore sequencing technology, resulting in a lack of nanopore HLA genotyping tools that can be truly used in clinical practice.

[0003] At present, most of the methods for HLA typing using nanopore sequencing use existing second-generation sequencing software, such as NGSEngine, SeqNext-HLA, HLA-LA, Athlon2, etc., or methods built by researchers based on typing principles for internal use only. These software or methods are generally not optimized based on the characteristics of nanopore sequencing data, resulting in the performance of HLA typing using nanopore sequencing data being difficult to meet the clinical requirement of 100% typing accuracy for HLA quartiles. Summary of the invention

[0004] In view of this, the present invention provides a gene haplotype typing method, device and medium based on sequencing data to solve the above problems.

[0005] In order to solve the above technical problems, the present invention provides a gene haplotype typing method based on sequencing data, comprising:

[0006] Acquiring sequencing data of the gene to be tested from the sample to be tested;

[0007] Sequencing data is split into genes according to the reference sequence, and then variant information detection and variant information filtering are performed; sequence splitting is performed based on the sequence after variant information filtering;

[0008] Among them, sequence splitting includes: determining the number of variant sites in the sequence:

[0009] If it is 0, it is considered homozygous and no split is performed;

[0010] If it is 1, the haplotype sequence is directly split based on the variant position;

[0011] If it is greater than 1, the sequencing sequences covering the variant position are first counted, and all the sequences covered at the same time are traversed by group, and the variant base combination of the same group in each sequence is recorded; based on the variant base combination, the sequence is sorted from high to low according to the number of types, and the variant base combination with the largest number in the same group is calculated, and the haplotype sequence is retained and split according to the proportion of different variant base combinations in all combinations;

[0012] The haplotype sequence after sequence splitting is corrected to obtain the haplotype typing result of the gene to be tested of the sample to be tested.

[0013] As an optional manner, the sequencing is nanopore sequencing, and / or the gene to be tested is an HLA gene, and / or the sequencing data does not include sequences whose data quality does not meet preset expectations.

[0014] As an optional method, before detecting the variation information, sequence shearing is also included to remove the part of the sequencing sequence that does not exist in the reference sequence, which includes:

[0015] Based on the sequencing sequence information belonging to each gene, the second reference sequence is compared in each gene range class, and typing calculation is performed to complete gene sequence shearing; wherein the second reference sequence is a reference sequence that does not contain UTR information;

[0016] The classification calculation includes:

[0017] The sequence of any gene is compared with the reference sequence of the gene in the second reference sequence and then filtered. Each sequence is parsed and cut through the comparison result file.

[0018] As an optional method, variant information detection includes:

[0019] The sequencing sequence is compared with the third reference sequence and filtered to generate variant detection statistics; wherein the third reference sequence is the first reference sequence of each gene selected from the reference sequence that does not contain UTR information;

[0020] It includes:

[0021] Set the comparison matching symbols, which include match, mismatch, base insertion and base deletion; perform comparison analysis based on the indicator attributes in the comparison result file, and count the number of each matching symbol for each position in the unique reference sequence of a single gene; retain the different bases and corresponding quantity information in the statistical match, mismatch, base insertion and base deletion results to the corresponding variation detection result file, and generate variation detection statistics;

[0022] Mutation information filtering is used to filter the results of mutation detection statistics, including calculation of mutation site coverage depth, site mutation base frequency and detection mutation recalculation, and filtering using the following methods:

[0023] Pre-set the threshold of site coverage depth, the threshold of site variant base frequency, and the threshold of the ratio of the major and minor alleles of the positive chain to the major and minor alleles of the negative chain; remove the variant sites that do not meet the threshold; wherein, the detection variant recalculation includes recalculating the ratio of the major and minor alleles of the positive chain to the major and minor alleles of the negative chain for the obtained variant sites;

[0024] Among them, when recalculating, the variant sites in the exon interval are preferentially selected; if no variant site is detected in the exon interval, the interval is further expanded to the gene intron region for variant site screening.

[0025] As an optional method, for sequences with a number of variant sites greater than 1, the sequencing sequences covering the variant sites are counted, and after the sites are arranged in order based on the number of sites of the variant information obtained, each site is taken as a reference site in turn according to the variant site information obtained by the detection, and the association information between the reference site and each other site in the real sequencing sequence is calculated; after all the calculations are completed, the association information of any reference site with all other sites is grouped to form a group of site association information;

[0026] The results of calculating each benchmark site are ranked by the number of associated sites in the group, and the site combination with the largest ranking is selected as a candidate for haplotype splitting to obtain the first candidate haplotype and the second candidate haplotype sequence;

[0027] The first candidate haplotype and the second candidate haplotype sequences are filtered to obtain the first haplotype sequence and the second haplotype sequence.

[0028] As an optional method, any reference site and the calculated site together account for more than 60% of the total base number of the site, and each of the two accounts for more than 12%.

[0029] As an optional manner, filtering the first candidate haplotype and the second candidate haplotype sequences to obtain the first haplotype and the second haplotype sequences includes:

[0030] According to the first candidate haplotype, the second candidate haplotype, and the base combination of the first candidate haplotype and the second candidate haplotype, each comparison sequence is filtered one by one, and the sequences whose haplotype site consistency is greater than the preset value are divided into the corresponding haplotype until the first haplotype sequence and the second haplotype sequence are split.

[0031] As an optional method, the haplotype sequence after sequence splitting is corrected including:

[0032] Sort the sequences belonging to the first haplotype from long to short, select the total sequences from long to short according to a preset number, use the preset minimum value as the template sequence, and extract the longest preset number of sequences outside the template sequence from the remaining sequences;

[0033] Taking each template sequence as a reference in turn, all sequences extracted for correction are aligned to the template sequence at the same time and then corrected. After correction, high-quality sequences of the first haplotype and the second haplotype, as well as homozygous high-quality sequences, are obtained.

[0034] As an optional method, the haplotype sequence after sequence splitting is corrected to obtain the haplotype typing result of the gene to be tested of the sample to be tested, including comparing the haplotype sequence with the reference gene sequence, performing typing scoring and verification;

[0035] It includes:

[0036] The high-quality haplotype sequence results were compared with the second reference sequence, and all the comparison results were filtered and scored as preliminary candidate typing results;

[0037] The typing results are verified and screened by comparing the differences in the exon region between the candidate typing and the high-quality monogamous sequence obtained after actual sequencing analysis; the matching score between the typing sequence of each gene and the high-quality monogamous sequence is obtained;

[0038] Based on the matching score, the variant sites of each typing sequence are back-scored, and the scoring results are sorted to obtain the final typing results of a single gene.

[0039] As an optional method, scoring of preliminary candidate typing includes:

[0040] The homozygous high-quality sequence is compared with the second reference sequence, and all the comparison results are used as preliminary candidate typing. Based on the comparison results, each preliminary candidate typing is scored and counted according to the following statistical method:

[0041] Set the reference length and alignment match symbols, which include match, mismatch, base insertion and base deletion; score the preliminary candidate typing marked with mismatch, insertion and deletion match symbols; and sort from high to low according to the scoring results.

[0042] As an optional method, verification and screening of typing results include:

[0043] Select the highest-scoring genotype as the common reference sequence for all candidate genotypes, and align all preliminary genotype sequences, including the highest-scoring genotype, to the selected common reference sequence;

[0044] Use a loop to traverse the public sequence. If there is any mismatch between the public reference sequence of any site and the comparison results of other candidate typing, the position of the site and the differences of all candidate typing at that position are marked for subsequent comparison with high-quality sequences.

[0045] The high-quality sequence is aligned to the public reference sequence, and then each position where there is any mismatch between other candidate typing and the public reference typing sequence in the previous step is traced back. Whenever the position is traced back, the base of the high-quality sequence aligned to the position is compared with all typing one by one. If the candidate typing and the high-quality sequence base are exactly the same, the typing score is increased; if not, it is not increased, and the scoring result is recorded;

[0046] After completing the scoring, sort them from high to low according to the scores.

[0047] As an optional method, if there is a mismatch in any exon region, a new candidate that fully matches the mismatched exon region is further identified from the database and matched and scored with the original candidate again. After the comparison is completed, the scores are sorted from high to low to obtain the final typing result of the single gene. The highest ranking is the final typing result. If there is a mismatch between the high-quality sequence and the final typing, a mismatch prompt message will be given.

[0048] As an optional method, after typing and verification, before obtaining the haplotype typing result of the gene to be tested of the sample to be tested, the following steps are also included: integrating the scoring and verification results, and the integration includes:

[0049] The final typing results of each gene are integrated, and the target region variation calculation is performed on the obtained relevant typing of the gene to be tested to obtain the final typing results of the gene to be tested for a single sample to be tested.

[0050] As an optional method, mutation calculation includes:

[0051] Any high-quality typing sequence and the final typing result are compared with the reference sequence. According to the compared reference sequence, the comparison with the reference sequence is extracted through the annotation information of the target region of the reference sequence. If there is an inconsistent part between the high-quality sequence and the reference sequence in the preset gene and / or an inconsistent part between the high-quality sequence and the reference sequence in another preset gene, it is recorded as a key site to indicate that there is a variation in the core exon; if there is a variation in other exons of the two types of genes, it is recorded as a common site, indicating that there is a variation in other exons.

[0052] On the other hand, the present invention also provides an electronic device, comprising: a memory for storing a computer program; and a processor for implementing the steps of the above-mentioned gene haplotype typing method based on sequencing data when executing the computer program.

[0053] On the other hand, the present invention also provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps of the genetic haplotype typing method based on sequencing data as described above are implemented.

[0054] The beneficial effects of the present invention are:

[0055] The present invention 1. simplifies complex calculations by first splitting the sequence into genes, avoiding multiple repeated calculations between highly similar sequences; 2. filters the variation detection information by cutting the sequence on one side of the variation site, amplifying the local information of the site, and reducing the interference of sequencing noise, which can better be compatible with nanopore data and reduce the impact of noise on variation detection; extracts haplotype mutation information by correlating and merging known sequence variations in pairs, which does not rely on prior experience or training models, can automatically adapt to the current actual sequencing results, effectively extracts haplotype mutation feature information, and has a flexible splitting method, allowing a certain fault tolerance space for splitting, which has a good effect on nanopore data; 3. obtains high-quality haplotype sequence by first splitting the haplotype and then correcting the sequence 1. The sequence is not completely dependent on the sequence alignment results. The final evaluation of similar typing is done by combining alignment with back-tracking variation information scoring. The typing difference information can be refined to reduce the interference of sequencing data noise on the results. The haplotype results are verified multiple times by initial typing plus correction to make the results more accurate. 2. The detection of core exon and exon interval variation information of actual sequencing sequences and known typing results can provide better readability and quantitative judgment indicators for the accuracy of the typing results, and at the same time, it can speculate on possible new HLA typing. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A schematic diagram of a gene haplotype typing method based on sequencing data provided by an embodiment of the present invention;

[0057] Figure 2 A simplified diagram of the generation of the haplotype consensus sequence provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0058] In order to enable those skilled in the art to better understand the technical solution of the present invention, the present invention is further described in detail below in conjunction with specific implementation methods.

[0059] See also Figure 1 This embodiment provides a gene haplotype typing method based on sequencing data, comprising:

[0060] Acquiring sequencing data of the gene to be tested from the sample to be tested;

[0061] Sequencing data is split into genes according to the reference sequence, and then variant information detection and variant information filtering are performed; sequence splitting is performed based on the sequence after variant information filtering;

[0062] Among them, sequence splitting includes: determining the number of variant sites in the sequence:

[0063] If it is 0, it is considered homozygous and no split is performed;

[0064] If it is 1, the haplotype sequence is directly split based on the variant position;

[0065] If it is greater than 1, the sequencing sequences covering the variant position are first counted, and all the sequences covered at the same time are traversed by group, and the variant base combination of the same group in each sequence is recorded; based on the variant base combination, the sequence is sorted from high to low according to the number of types, and the variant base combination with the largest number in the same group is calculated, and the haplotype sequence is retained and split according to the proportion of different variant base combinations in all combinations;

[0066] The haplotype sequence after sequence splitting is corrected to obtain the haplotype typing result of the gene to be tested of the sample to be tested.

[0067] As an optional method, the sequencing data is split into genes according to the reference sequence to divide the aligned sequencing sequences into the best aligned genes one by one, including: comparing and analyzing the acquired sequencing data with the first reference sequence to generate an alignment file; filtering the alignment file to remove unintentional amplification sequences and / or chimeric sequences and then parsing; identifying the corresponding reference sequence to which each sequencing sequence is aligned, and the gene information to which the reference sequence belongs, and dividing the aligned sequencing sequences into the best aligned genes one by one; wherein the first reference sequence is a gene representative sequence used to determine the gene from which the sequence comes.

[0068] In addition, before detecting the variation information, this embodiment also includes sequence shearing to remove the portion of the sequencing sequence that does not exist in the reference sequence, which includes:

[0069] Based on the sequencing sequence information belonging to each gene, the second reference sequence is compared in each gene range class, and typing calculation is performed to complete gene sequence shearing; wherein the second reference sequence is a reference sequence that does not contain UTR information; wherein the typing calculation includes:

[0070] The sequence of any gene is compared with the reference sequence of the gene in the second reference sequence and then filtered. Each sequence is parsed and cut through the comparison result file.

[0071] The above-mentioned variant information detection is to compare and filter the sequencing sequence with the third reference sequence to generate variant detection statistics; wherein the third reference sequence is the first reference sequence of each gene selected from the reference sequence that does not contain UTR information; specifically, the sampling is performed as follows:

[0072] Set the comparison matching symbols, which include match, mismatch, base insertion and base deletion; perform comparison analysis based on the indicator attributes in the comparison result file, and count the number of each matching symbol for each position in the unique reference sequence of a single gene; retain the different bases and corresponding quantity information in the statistical match, mismatch, base insertion and base deletion results to the corresponding variation detection result file, and generate variation detection statistics;

[0073] Mutation information filtering is used to filter the results of mutation detection statistics, including calculation of mutation site coverage depth, site mutation base frequency and detection mutation recalculation, and filtering using the following methods:

[0074] Pre-set the threshold of site coverage depth, the threshold of site variant base frequency, and the threshold of the ratio of the major and minor alleles of the positive chain to the major and minor alleles of the negative chain; remove the variant sites that do not meet the threshold; wherein, the detection variant recalculation includes recalculating the ratio of the major and minor alleles of the positive chain to the major and minor alleles of the negative chain for the obtained variant sites; wherein, when performing the recalculation, the variant sites in the exon interval are preferentially selected; if no variant site is detected in the exon interval, the interval is further expanded to the intron region of the gene for variant site screening.

[0075] Based on the above-mentioned shearing, detection and filtering results, further, in this embodiment, for sequences with a number of variant sites greater than 1, the sequencing sequences covering the variant sites are counted, and after the number of sites based on the obtained variant information is arranged in order, each site is taken as a reference site in turn according to the variant site information obtained by the detection, and the association information between the reference site and each of the remaining sites in the real sequencing sequence is calculated; after all the calculations, the association information between any reference site and all other sites is grouped to form a group of site association information; the number of associated sites in the group calculated for each reference site is sorted, and the site combination with the largest number of rankings is selected as a candidate for haplotype splitting to obtain a first candidate haplotype and a second candidate haplotype sequence; the first candidate haplotype and the second candidate haplotype sequences are filtered, and each comparison sequence is filtered one by one according to the first candidate haplotype, the second candidate haplotype, and the base combination of the first candidate haplotype and the second candidate haplotype, and the sequences whose haplotype site consistency is greater than a preset value are divided into corresponding haplotypes until the first haplotype sequence and the second haplotype sequence are split.

[0076] The haplotype sequence obtained at this time is not a high-quality sequence that can be used directly, so it needs to be corrected. Sort the sequences belonging to the first haplotype from long to short, select the total sequence from long to short according to the preset number, use the preset minimum value as the template sequence, and extract the longest preset number of sequences outside the template sequence from the remaining sequences; take each template sequence as a reference, and align all the sequences extracted for correction to the template sequence at the same time and then correct them. After correction, high-quality sequences of the first haplotype and the second haplotype, as well as homozygous high-quality sequences, are obtained.

[0077] Afterwards, the high-quality haplotype sequence results are compared with the second reference sequence, and all the comparison results are used as preliminary candidate typing and filtered and scored; the typing results are verified and screened by comparing the differences in the exon region between the candidate typing and the high-quality monogamous sequence obtained after actual sequencing analysis; the matching score between the typing sequence of each gene and the high-quality monogamous sequence is obtained; based on the matching score, the variant sites of each typing sequence are back-scored, and the final typing results of a single gene are obtained after sorting the scoring results.

[0078] Among them, scoring the preliminary candidate typing includes:

[0079] The homozygous high-quality sequence is compared with the second reference sequence, and all the comparison results are used as preliminary candidate typing; based on the comparison results, each preliminary candidate typing is scored and counted according to the following statistical method: the reference length and the comparison matching symbol are set, and the comparison matching symbol includes match, mismatch, base insertion and base deletion; the preliminary candidate typing marked with mismatch, insertion and deletion matching symbols is scored; and the scoring results are sorted from high to low.

[0080] Verification and screening of typing results include:

[0081] Select the highest-scoring genotype as the common reference sequence for all candidate genotypes, and align all preliminary genotype sequences, including the highest-scoring genotype, to the selected common reference sequence;

[0082] Use a loop to traverse the public sequence. If there is any mismatch between the public reference sequence of any site and the comparison results of other candidate typing, the position of the site and the differences of all candidate typing at that position are marked for subsequent comparison with high-quality sequences.

[0083] The high-quality sequence is aligned to the public reference sequence, and then each position where there is any mismatch between other candidate typing and the public reference typing sequence in the previous step is traced back. Whenever the position is traced back, the base of the high-quality sequence aligned to the position is compared with all typing one by one. If the candidate typing and the high-quality sequence base are exactly the same, the scoring score of the typing is increased; if not, it is not increased, and the scoring results are recorded; after the scoring is completed, it is sorted from high to low according to the score.

[0084] As an optional method, if there is a mismatch in any exon region, a new candidate that fully matches the mismatched exon region is further identified from the database and matched and scored with the original candidate again. After the comparison is completed, the scores are sorted from high to low to obtain the final typing result of the single gene. The highest ranking is the final typing result. If there is a mismatch between the high-quality sequence and the final typing, a mismatch prompt message will be given.

[0085] Furthermore, this embodiment also performs processing before obtaining the haplotype typing result of the gene to be tested of the sample to be tested after low typing scoring and verification, integrates the final typing result of each gene, performs target region variation calculation on the obtained typing related to the gene to be tested, and obtains the final typing result of the gene to be tested of a single sample to be tested. The variation calculation process is to compare any typing high-quality sequence and the final typing result with the reference sequence, and extract the comparison with the reference sequence according to the compared reference sequence and the annotation information of the target region of the reference sequence, if there is a part of the high-quality sequence in the preset gene that is inconsistent with the reference sequence and / or a part of the high-quality sequence in another preset gene that is inconsistent with the reference sequence, it is recorded as a key site to indicate that there is a variation in the core exon; if there is a variation in other exons of the two types of genes, it is recorded as a common site, indicating that there is a variation in other exons.

[0086] Based on the above steps, this embodiment discloses a high-performance gene haplotype typing method based on sequencing data. Furthermore, in this embodiment, the sequencing method selected when describing the scheme is nanopore sequencing, and the gene to be tested is selected as HLA gene, so as to supplement the HLA typing method based on nanopore sequencing data and promote the HLA typing method or product based on nanopore sequencing to be applied to clinical practice more quickly.

[0087] In order to further illustrate the above solution, this embodiment provides a specific operation method, which is implemented according to the following steps:

[0088] S1. Obtain sequencing data. By adopting an amplicon experimental scheme, the sample to be tested is sequenced using a nanopore sequencing platform to obtain sequencing data of HLA-related genes. Optionally, the HLA-related genes involved in this embodiment include: HLA-I class: HLA-A, HLA-B, HLA-C; HLA-II class: HLA-DRB1, HLA-DRB3, HLA-DRB4, HLA-DRB5, HLA-DPA1, HLA-DPB1, HLA-DQ A1, HLA-DQB1, one or more genes.

[0089] S2. The sequencing data used in this embodiment does not include sequences whose data quality does not meet preset expectations. Therefore, the quality control of the sequencing data includes performing data quality control processing on the nanopore sequencing data obtained above to remove sequences with low data quality. In one implementation scenario, it can be: taking the base quality value Q>=7 as a threshold, filtering out sequences with an average base quality value lower than Q<7.

[0090] S3. Construction of typing database: As an optional method, you can download known HLA typing data information from the IMGT (https: / / www.ebi.ac.uk / ipd / imgt / hla / ) database. The downloaded database version number is IPD-IMGT / HLA3.54.0. Extract the gene sequences mentioned in S1 in the database, and organize them according to the relevant annotation information of the database to obtain the files required for subsequent analysis. Among them, the database organization method and the obtained files include:

[0091] a). Gene representative sequence (rREF) used to determine the source gene of the sequencing sequence: Use cd-hit software (version 4.8.1) to cluster the single genotyping sequences in the IMGT database by similarity, with a clustering threshold of 0.99, and then retain the representative of each cluster group (for alignment to the best gene in S4 below);

[0092] b) noUTR-REF sequence used for secondary alignment and preliminary typing identification: based on the original information of the IMGT gene sequence database, the utr information in the sequence was uniformly removed (based on S4, used for typing calculation in S5 below);

[0093] c) Reference sequences for gene variation detection and haplotype typing: From noUTR-REF, select the first reference sequence rnoUTR-REF of each gene in the database (used for gene variation information monitoring in S6 below);

[0094] d). Core exon sequences for non-full-length genotyping: exon sequences extracted based on the exon position information in the hla.xml file of the database, including the exon 2 sequences of HLA-I and HLA-II genes, and the exon 3 sequences of HLA-I genes (keyREF) (used for the output typing results of S12 below).

[0095] S4, Sequencing data split into genes: According to the sequencing data results after quality control obtained in S2, the minimap2 software is used for alignment analysis (optional parameters are: -a-xmap-ont--eqx), and the quality control data is aligned to the HLA gene representative full-length sequence database rREF constructed in S3, and the aligned sam format file is obtained. The samtool software is used to filter the alignment result file to remove unintentional amplification sequences or chimeric sequences. The filtering condition is flag tag 2308 (flag is the description of the sequence alignment in the sam file format), 2308 includes sequences that cannot be aligned, sequences that are not preferred for alignment, and sequences for supplementary alignment. The result file is then filtered for parsing, and the aligned sequencing sequences are divided into the best alignment genes one by one by identifying the corresponding reference sequence in the rREF to which each sequencing sequence is aligned, and the gene information to which the rREF reference sequence belongs in the IMGT database.

[0096] S5. Gene sequence shearing: Based on the sequencing sequence information belonging to each gene obtained in S4, the reference sequence of the gene corresponding to noUTR-REF is aligned in each gene range class, and typing calculation is performed.

[0097] Taking gene A as an example, the sequence of gene A is first aligned to the reference sequence of A in the gene sequence noUTR-REF for typing obtained in S3, and then the alignment result is filtered with reference to flag 2308 in S4. Each sequence is aligned and parsed through the cigar (CompactIdiosyncratic GappedAlignmentReport) string in the alignment result file (the cigar string is in the sam format file, and the base matching method in the alignment result is recorded using different characters). The sequence part indicating clipalignment in the sequence (the sequence is not completely matched with the reference, and only one segment has a matching alignment result) is cut off. This part of the sequence indicates the part that does not exist in the reference sequence, where "H" represents Hardclip in the alignment, and "S" represents Softclip in the alignment (the sequence has only one segment matching with the reference, and the segment that cannot be aligned is recorded in cigar as a number plus "S").

[0098] S6, gene variation information detection: the sequence retained after clipalignment in S5 is aligned to the reference sequence rnoUTR-REF obtained in S3 for gene variation detection and haplotype typing, and the alignment result is filtered with reference to flag 2308 in S4. Each sequence is aligned and parsed through the cigarstring in the file. In the cigarstring of S6, the "=" symbol indicates a match, the "X" symbol indicates an incorrect match, the "I" symbol indicates that the sequenced sequence has a base insertion compared to the reference sequence, and the "D" symbol indicates that the sequenced sequence has a base deletion compared to the reference sequence. By analyzing the above features, the total number of each symbol is calculated for each position in the unique reference sequence of a single gene. Assuming that the 10th base in the reference sequence is "A", there are 100 "=" symbols in the 200 comparison results, indicating that there are 100 sequences identical to the reference base at this position, and 100 "X" symbols, indicating that there are 100 base mismatches with the reference sequence, that is, there is a mutation SNP (single nucleotide polymorphism). Assuming that in the comparison sequence corresponding to the "X" symbol, the base information at this position is all "G", it means that the base at this position is composed of 100 "A" and 100 "G", which may be a heterozygous mutation from base "A" to base "G". The different bases and corresponding quantity information (coverage depth) obtained from this part of the statistics are retained in the corresponding mutation detection result file (mutation_file). For other types of mutations, such as insertions or deletions, they are all recorded and retained in this way.

[0099] S7, mutation information filtering: Filter the mutations (SNPs) detected according to the mutation detection statistics (mutation_file) in S6 to obtain reliable SNP results. The filtering conditions are to remove sites that do not meet the following three points:

[0100] a). The site coverage depth is required to be more than 100X, that is, the number of sequences covering the site is greater than or equal to 100;

[0101] b). The mutation base frequency of the site is required to be greater than or equal to 0.15;

[0102] c). Recalculate the proportion of the positive chain major and minor alleles and the negative chain major and minor alleles for the obtained SNPs, and the difference is required not to exceed 0.14. When recalculating, give priority to SNPs in the exon interval. If no SNP is detected in the exon interval, continue to expand the interval to the gene intron region for SNP screening. The method for recalculating the allele proportion is as follows:

[0103] 8bp and 300bp are taken from the upstream and downstream of the SNP location to construct a new reference sequence (sSNP-REF). The value is flexibly adjusted according to the location of the SNP. The adjustment method is: if the downstream meets the 300bp length, 8bp upstream and 300bp downstream are preferred. If the downstream is less than 300bp in length, the sequence is taken upstream; after aligning the sequencing sequence to the newly constructed reference sequence (sSNP-REF), the alignment base information of the SNP, the proportion of the positive chain main and minor alleles and the negative chain main and minor alleles are counted. The above data processing is used to eliminate the misdetected or missed SNP sites caused by the interference caused by the alignment offset when the data quality value is not high.

[0104] S8, haplotype sequence splitting: According to the filtered variation detection results obtained in S7, the haploid sequence is split based on the position information of the sites where the variation exists. The specific examples are as follows:

[0105] a). If the number of SNPs after filtering is 0, it is considered homozygous and no splitting is performed. Go directly to S9 for analysis.

[0106] b). If the number of SNPs after filtering is 1, indicating that the two haplotypes differ by only one base, the sequence is directly split according to the base information of the mutation position. The splitting process is: if the sequence seq1 is typed as "A" at the SNP, the sequence is classified as the first haplotype sequence (first haplotype sequence), and is subsequently referred to as the first haplotype (first haplotype); if the sequence seq1 is typed as "G" at the SNP, the sequence is classified as the second haplotype sequence (second haplotype sequence), and is subsequently referred to as the second haplotype (second haplotype);

[0107] c). If the number of SNPs after filtering is greater than 1, indicating that there are multiple base differences between the two haplotypes, the sequencing sequences covering the mutation sites are first counted. Assuming that S7 obtains 10 mutation sites, the 10 sites are represented by numbers 1 to 10, and the association calculation is performed with two sites as a combination. During the calculation, all sequences covering the two sites at the same time are traversed, and the base combination of the two sites in each sequence is recorded. Assuming that the currently calculated sites are site 1 and site 2, the total number of sequences covering the two sites at the same time is 200, of which the number of sequences with site 1 being "A" and site 2 being "G" is 36 (accounting for 18%), and the number of sequences with site 1 being "C" is 18%. There are 130 sequences with "T" at position 2 (accounting for 65%), and the remaining 34 sequences are "T" at position 1, "A" at position 2, or "-" at position 1 ("-" indicates that the position is a base deletion), and "A" at position 2, etc.; then these base combinations are sorted from high to low according to the number, and the two base combinations with the largest number are calculated. It is required that the sum of the number of sequences of the two accounts for more than 60% of the total number of sequences, and the proportion of the second base combination in the total combination is greater than 12%. In this way, the two base combinations with the largest number are retained, and it is considered that there is a correlation between position 1 and position 2. The relationship between the other positions will also be calculated one by one without duplication;

[0108] d). According to the above association calculation results, the positions with association relationships are merged. Assuming that the relationship between site 1 and site 2 in a is satisfied, in other calculation results, site 1 and site 5, site 1 and site 7, and site 1 and site 8 are all associated. Finally, sites 1, 2, 5, 7, and 8 are retained for further calculation. If sites 3, 4, 6, 9, and 10 have no association relationship with other sites other than their own sites, it is considered that the site is not related to any other mutation site and does not participate in subsequent calculations. When merging, a single base of one of the sites is used as an index, such as When the base at site 1 is "C", the base combination at sites 2, 5, 7, and 8 is considered as a potential haplotype combination coexisting with site 1 (hereinafter referred to as the first candidate haplotype), and when the base at site 1 is "A", the base combination at sites 2, 5, 7, and 8 is considered as another potential haplotype combination coexisting with site 1 (hereinafter referred to as the second candidate haplotype). After integration, the longest candidate haplotype is selected from the first candidate haplotype as the first haplotype, and the haplotype corresponding to the first haplotype and containing the same site is selected from the second candidate haplotype as the second haplotype. The bases of the selected haplotypes at each candidate site are mutually exclusive. For example, site 1 is base "A" in the first haplotype and base "C" in the second haplotype. If a site is the same base in the first haplotype or the second haplotype, the site is removed from both haplotype sites.

[0109] e). According to the first haplotype and the second haplotype obtained in b and the combination of their bases, the compared sequences are split into haplotypes. If the two haplotypes have 5 sites within 1, 2, 5, 7, and 8, each compared sequence is filtered one by one. For sequences with a consistency of more than 80% for each haplotype site, the sequence is divided into the corresponding haplotype. For example, if the sequence seq1 meets the coverage of sites 1, 2, 5, and 7, and has bases at these 4 sites, and these 4 bases are consistent with the base characteristics of the first haplotype, the sequence is judged to be a sequence derived from the first haplotype. According to the description in this section, the comparison results are split into sequences of the first haplotype and the second haplotype.

[0110] S9. Haplotype sequence correction: According to the haplotype sequence results obtained by S8 splitting, sequence correction is performed to obtain high-quality haplotype sequences. Taking the first haplotype as an example, during correction, the sequences belonging to the first haplotype are first sorted from long to short, and 15% of the total sequences are selected from long to short, but the number of sequences is limited to a maximum of 30 and a minimum of 10 as template sequences. Subsequently, the longest 80 sequences outside the template sequences are extracted from the remaining sequences to correct these template sequences. A maximum of 80 sequences are extracted. During correction, the software racon (version 1.5.0) is used. Each template sequence is used as a reference in turn. All sequences extracted for correction are simultaneously aligned to the template sequence and then corrected. After correction, high-quality sequences of the first haplotype and the second haplotype, or homozygous high-quality sequences (hqSEQ) are obtained.

[0111] S10, high-quality haplotype sequence typing: According to the haplotype high-quality sequence results obtained in S9, the high-quality sequence (hqSEQ) was aligned to the reference gene sequence (noUTR_REF) with both ends of the utr removed constructed in S3, and all the alignment results were used as preliminary candidate typing. The alignment was performed using minimap2 software (parameters: -a-xmap-ont--eqx), and the alignment results were filtered according to the flag2308 threshold described in S4. According to the alignment results, each preliminary candidate typing was scored and counted according to the following statistical method, where the reference sequence length was recorded as reflength, cigarstring contained sequence alignment information, where "=" was recorded as a match, and its sum was represented by M, "X" was recorded as a mismatch, and its sum was represented by N, "I" was recorded as an inserted base compared to the aligned reference sequence, and its sum was represented by I, and "D" was recorded as a missing base compared to the aligned reference sequence, and its sum was represented by D. The total alignment length is calculated as the sum of the total number of matches and the total number of mismatches, and mismatches, insertions, and deletions are penalized. When penalizing, mismatches, insertions, or deletions are penalized 1 to distinguish highly similar sequences, and the sequences are sorted from high to low according to the score values.

[0112] S11. Verification of typing results: Based on the preliminary typing results of S10, the typing results were verified and screened by comparing the differences in the exon region between the candidate typing and the hqSeq sequences obtained after actual sequencing. The highest-scoring typing was selected as the common reference sequence (tREF) for all candidate typings, and all preliminary typing sequences including the highest-scoring typing were aligned to the selected common reference sequence (tREF). The alignment software and parameters were the same as those of S10. Pysam (version 0.21.0) was used to parse the alignment results, and the common sequence was traversed in a loop. If there was any mismatch between the common reference sequence of a certain site and the alignment results of other candidate typings, the position of the site and the differences of all candidate typings at that position were marked for subsequent comparison with the high-quality sequence (hSEQ) alignment results. The high-quality sequence (hqSEQ) is aligned to the public reference sequence (tREF), and then each position where there is any mismatch between other candidate typing and the public reference typing sequence (tREF) in the previous step is traced back. Whenever the position is traced back, the base of the high-quality sequence aligned to the position is compared with all typing one by one. If the candidate typing is exactly the same as the high-quality sequence base, the typing score is increased by 10. If not, it is not increased. The scoring results are recorded and all sites with specific scores for each typing are reflected in the subsequent results, as well as sites with inconsistent scores. After the scoring is completed, the scores are sorted from high to low. The higher the score, the higher the consistency between the sequencing result and the typing result.

[0113] S12. Obtain the final typing results: Since some genes in the IMGT database do not have full-length gene reference sequences, and some even only contain information on a single exon sequence, in order to avoid typing misjudgments caused by such reasons, the high-quality sequence (hqSEQ) obtained in S9 is used. For the HLA-I and HLA-II class genes described in S3, the hqSEQ is aligned to the exon reference sequence (keyREF) constructed in S3. Take HLA-II class genes as an example: the comparison results are compared using the method in S11 to perform SNP backtracking and scoring, and the scores are sorted from high to low, recorded as the core exon typing results, and then the preliminary typing obtained in S11 is compared. If the preliminary typing result and the best score of the exon scoring result are in the same Ggroup, and the same Ggroup means that the two are consistent in the core exon region, then the preliminary typing result obtained in S11 is used as the final typing result; if the preliminary typing result and the best typing of the exon scoring result are in different groups, then the preliminary typing is compared with the highest score of the exon typing result, and the comparison S is consistent with the score S in S11. After the comparison is completed, the scores are sorted from high to low to obtain the final typing result of a single gene.

[0114] S13, Typing result integration: According to the single genotyping results obtained in S12, the exon region mutation calculation is performed on the obtained HLA gene-related typing. Each gene is calculated separately. Taking gene A as an example, the typing results related to gene A, the first haplotype and the second haplotype or homozygous typing, firstly, the high-quality sequence (hqSEQ) obtained in S9 corresponding to the typing is aligned to the reference sequence in the IMGT database corresponding to the typing result obtained in S12. According to the aligned reference sequence, the annotation information of the exon interval in the IMGT database is used to extract the sequence aligned to the reference sequence. If there is a part where the high-quality sequence and the reference sequence are inconsistent in exon 2 or 3 of HLA class I gene, or there is a part where the high-quality sequence and the reference sequence are inconsistent in exon 2 of HLA class II gene, it is recorded as KeySNP, indicating that there is a mutation in the core exon. If there is a mutation in other exons of the two types of genes, it is recorded as SNP, indicating that there is a mutation in other exons. After completing the above results, the typing results of each gene (different genes in the HLA gene type, including 11 types in this embodiment) are integrated to obtain the final typing results of 11 HLA genes in a single sample.

[0115] See also Figure 2 Through the above scheme, this embodiment simplifies the complex calculation by splitting the sequence into genes first. When splitting the gene, the sequence is compared to the simplified representative sequence database described in S3, avoiding multiple repeated calculations between highly similar sequences. Taking the version IMGT (3.54.0) database as an example, the reference sequence of the target gene is reduced from 18734 to 138, reducing more than 99% of redundant calculations, greatly reducing the running resource consumption and running time consumption of the program. Among them, Figure 2 Unphasedreads indicates the sequencing nanopore sequence that has not completed haplotype splitting; Sequencingerror indicates the random error in the sequencing sequence due to sequencing reasons; Phasedreads indicates the sequence that has completed haplotype splitting; Haplotype 1 indicates the first haplotype sequence after haplotype splitting; Haplotype2 indicates the second haplotype sequence after haplotype splitting; Consensus indicates the high-quality sequence obtained by self-correction of the first haplotype or the second haplotype sequence.

[0116] This embodiment uses a local zoom method (by cutting the sequence on one side of the variation site, amplifying the local information of the site, and reducing sequencing noise interference) to filter the variation detection information, which can better be compatible with nanopore data and reduce the impact of noise on variation detection; the haplotype mutation information is extracted by correlating and merging known sequence variations. This method does not rely on prior experience or training models, can automatically adapt to the current actual sequencing results, and effectively extracts the mutation feature information of the haplotype. At the same time, the splitting method is flexible, allowing a certain fault tolerance space for splitting, and has a good effect on nanopore data.

[0117] Moreover, in some specific scenarios, the method of first splitting haplotypes and then correcting the sequences is adopted to obtain high-quality haplotype sequences. This ensures that the correction of the first haplotype sequence will not be interfered by other haplotype data during the correction process, and the results are more accurate and reliable. Especially when the ratio of two haplotypes deviates seriously from 1:1, the haplotype sequence information with a lower proportion can be more completely retained.

[0118] When obtaining the final typing results, this embodiment creatively improves the scoring process compared to the prior art or industry knowledge, so that this embodiment does not completely rely on the sequence alignment results, and uses alignment combined with backtracking variation information scoring to perform a final evaluation of similar typing, which can refine the typing difference information and reduce the interference of sequencing data noise on the results; the haplotype results are verified multiple times by using the initial typing plus correction method, making the results more accurate.

[0119] In particular, in some special scenarios, the detection of the core exon and exon interval variation information of the actual sequencing sequence and the known typing results in this embodiment can provide better readability and quantitative judgment indicators for the accuracy of the typing of the test results, and at the same time, it can speculate on possible new HLA typing. Therefore, this embodiment is suitable for HLA typing based on full-length sequencing data and has better compatibility with nanopore data with higher noise.

[0120] On the other hand, this embodiment further provides an electronic device, comprising: a memory for storing a computer program; and a processor for implementing the steps of the above-mentioned gene haplotype typing method based on sequencing data when executing the computer program.

[0121] On the other hand, this embodiment further provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps of the gene haplotype typing method based on sequencing data as described above are implemented.

[0122] The above are only preferred embodiments of the present invention. It should be noted that the above preferred embodiments should not be regarded as limiting the present invention, and the protection scope of the present invention should be based on the scope defined by the claims. For ordinary technicians in this technical field, several improvements and modifications can be made without departing from the spirit and scope of the present invention, and these improvements and modifications should also be regarded as the protection scope of the present invention.

Claims

1. A gene haplotype typing method based on sequencing data, comprising: Acquiring sequencing data of the gene to be tested from the sample to be tested; Splitting the sequencing data into genes according to the reference sequence and then performing variation information detection and variation information filtering; Sequence splitting based on the sequence filtered by variation information; The sequence splitting includes: determining the number of variant sites in the sequence: If it is 0, it is considered homozygous and no split is performed; If it is 1, the haplotype sequence is directly split based on the variant position; If it is greater than 1, first count the sequencing sequences covering the variant position, traverse all the sequences covered at the same time by group, and record the variant base combination of the same group in each sequence; based on the variant base combination, sort from high to low according to the number of types, calculate the largest number of variant base combinations in the same group, and retain and split them according to the proportion of different variant base combinations in all combinations to obtain haplotype sequences; The haplotype sequence after sequence splitting is corrected to obtain the haplotype typing result of the gene to be tested of the sample to be tested.

2. A gene haplotype typing method based on sequencing data according to claim 1, characterized in that: The sequencing is nanopore sequencing, and / or the gene to be tested is an HLA gene, and / or the sequencing data does not include a sequence whose data quality does not meet preset expectations.

3. A gene haplotype typing method based on sequencing data according to claim 1 or 2, characterized in that: The step of splitting the sequencing data into genes according to the reference sequence is to divide the aligned sequencing sequences into the best aligned genes one by one, which includes: The acquired sequencing data is compared and analyzed with the first reference sequence to generate a comparison file; the comparison file is filtered to remove unintentional amplification sequences and / or chimeric sequences and then analyzed; the corresponding reference sequence to which each sequencing sequence is compared and the gene information to which the reference sequence belongs are identified, and the compared sequencing sequences are divided into the best comparison genes one by one; wherein the first reference sequence is a gene representative sequence used to determine the gene from which the sequence comes.

4. A gene haplotype typing method based on sequencing data according to claim 1, characterized in that: Before the variation information detection, sequence shearing is also included to remove the part of the sequencing sequence that does not exist in the reference sequence, which includes: Based on the sequencing sequence information belonging to each gene, the second reference sequence is compared in each gene range class, and typing calculation is performed to complete gene sequence shearing; wherein the second reference sequence is a reference sequence that does not contain UTR information; The typing calculation includes: The sequence of any gene is compared with the reference sequence of the gene in the second reference sequence and then filtered, and each sequence is parsed and cut through the comparison result file.

5. A gene haplotype typing method based on sequencing data according to claim 1, characterized in that: The variation information detection includes: Compare and filter the sequenced sequence with a third reference sequence to generate variation detection statistics; wherein the third reference sequence is the first reference sequence of each gene selected from the reference sequence that does not contain UTR information; It includes: Setting comparison matching symbols, wherein the comparison matching symbols include match, mismatch, base insertion and base deletion; performing comparison analysis based on the indicator attributes in the comparison result file, and performing quantity statistics of each matching symbol for each position in the single gene unique reference sequence; retaining different bases and corresponding quantity information in the statistically obtained match, mismatch, base insertion and base deletion results to the corresponding variation detection result file, and generating variation detection statistics; The variation information filtering is used to filter the variation information of the variation detection statistics, which includes performing variation site coverage depth calculation, site variation base frequency calculation and detection variation recalculation on the variation detection statistics, and filtering by the following method: Pre-set the threshold of site coverage depth, the threshold of site variant base frequency, and the threshold of the ratio of the major and minor alleles of the positive chain to the major and minor alleles of the negative chain; remove the variant sites that do not meet the threshold; wherein the detection variant recalculation includes recalculating the ratio of the major and minor alleles of the positive chain to the major and minor alleles of the negative chain for the obtained variant sites; When performing the recalculation, the variant sites in the exon interval are preferentially selected; if no variant sites are detected in the exon interval, the interval is further expanded to the gene intron region for variant site screening.

6. A gene haplotype typing method based on sequencing data according to claim 1, characterized in that: For sequences with more than 1 variant sites, the sequencing sequences covering the variant sites are counted, and the sites are arranged in order based on the number of variant sites obtained. Then, according to the variant site information obtained by the test, each site is used as the reference site in turn, and the association information between the reference site and each other site in the real sequencing sequence is calculated; After all calculations, the association information of any reference site with all other sites is grouped to form a set of site association information; The results of calculating each benchmark site are ranked by the number of associated sites in the group, and the site combination with the largest ranking is selected as a candidate for haplotype splitting to obtain the first candidate haplotype and the second candidate haplotype sequence; The first candidate haplotype and the second candidate haplotype sequences are filtered to obtain a first haplotype sequence and a second haplotype sequence.

7. A gene haplotype typing method based on sequencing data according to claim 6, characterized in that: Any of the reference sites and the calculated sites account for more than 60% of the total base number of the site, and each of the two accounts for more than 12%.

8. A gene haplotype typing method based on sequencing data according to claim 6 or 7, characterized in that: The filtering of the first candidate haplotype and the second candidate haplotype sequences to obtain the first haplotype and the second haplotype sequences comprises: According to the first candidate haplotype, the second candidate haplotype, and the base combination of the first candidate haplotype and the second candidate haplotype, each comparison sequence is filtered one by one, and the sequences whose haplotype site consistency is greater than a preset value are divided into corresponding haplotypes until the first haplotype sequence and the second haplotype sequence are split.

9. A method for gene haplotype typing based on sequencing data according to any one of claim 1, characterized in that: The correction of the haplotype sequence after sequence splitting comprises: Sort the sequences belonging to the first haplotype from long to short, select the total sequences from long to short according to a preset number, use the preset minimum value as the template sequence, and extract the longest preset number of sequences outside the template sequence from the remaining sequences; Taking each template sequence as a reference in turn, all sequences extracted for correction are aligned to the template sequence at the same time and then corrected. After correction, high-quality sequences of the first haplotype and the second haplotype, as well as homozygous high-quality sequences, are obtained.

10. A gene haplotype typing method based on sequencing data according to claim 9, characterized in that: The haplotype sequence after sequence splitting is corrected to obtain the haplotype typing result of the gene to be tested of the sample to be tested, including comparing the haplotype sequence with the reference gene sequence, performing typing scoring and verification; It includes: The high-quality haplotype sequence results were compared with the second reference sequence, and all the comparison results were filtered and scored as preliminary candidate typing results; The typing results are verified and screened by comparing the differences in the exon region between the candidate typing and the high-quality monogamous sequence obtained after actual sequencing analysis; the matching score between the typing sequence of each gene and the high-quality monogamous sequence is obtained; Based on the matching score, the variant sites of each typing sequence are back-scored, and the scoring results are sorted to obtain the final typing result of a single gene.

11. A gene haplotype typing method based on sequencing data according to claim 10, characterized in that: Scoring of preliminary candidate typing includes: The homozygous high-quality sequence is compared with the second reference sequence, and all the comparison results are used as preliminary candidate typing. Based on the comparison results, each preliminary candidate typing is scored and counted according to the following statistical method: A reference length and a comparison match symbol are set, wherein the comparison match symbol includes a match, a mismatch, a base insertion and a base deletion; a preliminary candidate typing marked with a mismatch, an insertion and a deletion match symbol is scored; and the typing is sorted from high to low according to the scoring results.

12. A gene haplotype typing method based on sequencing data according to claim 11, characterized in that: The verification and screening of typing results include: Select the highest-scoring genotype as the common reference sequence for all candidate genotypes, and align all preliminary genotype sequences, including the highest-scoring genotype, to the selected common reference sequence; Use a loop to traverse the public sequence. If there is any mismatch between the public reference sequence of any site and the comparison results of other candidate typing, the position of the site and the differences of all candidate typing at that position are marked for subsequent comparison with high-quality sequences. The high-quality sequence is aligned to the public reference sequence, and then each position where there is any mismatch between other candidate typing and the public reference typing sequence in the previous step is traced back. Whenever the position is traced back, the base of the high-quality sequence aligned to the position is compared with all typing one by one. If the candidate typing and the high-quality sequence base are exactly the same, the typing score is increased; if not, it is not increased, and the scoring result is recorded; After completing the scoring, sort them from high to low according to the scores.

13. A gene haplotype typing method based on sequencing data according to claim 12, characterized in that: If there is an incorrect match in any exon region, a new candidate that completely matches the mismatched exon region is further identified from the database and matched and scored with the original candidate again. After the comparison is completed, the candidates are sorted from high to low according to the scores to obtain the final typing result of the single gene. The highest ranking is the final typing result. If there is a mismatch between the high-quality sequence and the final typing, a mismatch prompt message will be given.

14. A gene haplotype typing method based on sequencing data according to claim 10, characterized in that: After typing scoring and verification, before obtaining the haplotype typing result of the gene to be tested of the sample to be tested, the following steps are also included: integrating the scoring and verification results, and the integration includes: The final typing results of each gene are integrated, and the variation of the target region (such as the exon region) is calculated for the obtained relevant typing of the gene to be tested, so as to obtain the final typing result of the gene to be tested for a single sample to be tested.

15. A gene haplotype typing method based on sequencing data according to claim 14, characterized in that: The variation calculation includes: Any high-quality typing sequence and the final typing result are compared with the reference sequence. According to the compared reference sequence, the comparison with the reference sequence is extracted through the annotation information of the target region of the reference sequence. If there is an inconsistent part between the high-quality sequence and the reference sequence in the preset gene and / or an inconsistent part between the high-quality sequence and the reference sequence in another preset gene, it is recorded as a key site to indicate that there is a variation in the core exon; if there is a variation in other exons of the two types of genes, it is recorded as a common site, indicating that there is a variation in other exons.

16. An electronic device, characterized in that: include: Memory for storing computer programs; A processor, configured to implement the steps of the gene haplotype typing method based on sequencing data as claimed in any one of claims 1 to 15 when executing the computer program.

17. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the gene haplotype typing method based on sequencing data as claimed in any one of claims 1 to 15 are implemented.

Images