Non-targeted region genotype filling method based on whole exome sequencing

Abstract

The present invention discloses a non-targeted region genotype filling method, system, device and computer-readable storage medium based on whole exome sequencing. The method comprises: obtaining whole exome sequencing data of a target queue and a reference whole genome sequencing data set; filtering sites in the reference whole genome sequencing data set, and outputting SNP site information of the reference whole genome sequencing data set; calculating the GL of the SNP site of each sample based on the SNP site information and the whole exome sequencing data set, and obtaining the GL of each sample; merging the GL of each sample to obtain the GL of all samples; performing fragmentation processing on the reference whole genome sequencing data set to obtain genome fragment information; based on the GL of all samples, the SNP site information and the genome fragment information, using a machine learning algorithm to estimate the genotype of a non-targeted region of a single sample in the target queue, and obtaining a genotype estimation result of the non-targeted region of the single sample.

Description

Technical Field

[0001] The present invention relates to the technical field of gene prediction, and more specifically, to a non-targeted region genotype filling method based on whole exome sequencing and a system thereof. Background Art

[0002] With the rapid development of high-throughput sequencing technology (Next generation sequencing, NGS), more and more GWAS are trying to use sequencing technology for genotyping. Whole genome sequencing (WGS) can effectively cover most genetic mutation sites, but its disadvantage is that it is expensive and cannot be applied to large-scale cohort samples. In addition, if the sequencing depth is reduced for cost reasons, it is impossible to accurately identify the genotype status of low-frequency variants, which are often located in protein-coding regions and have important biological functions.

[0003] Whole Exome sequencing (WES) has been widely used to study the role of protein coding variation in genetic diseases. However, more than 98% of the sequences in the genome are non-coding sequences, and there are many important regulatory elements that can significantly affect gene expression. A large number of studies have shown that non-coding variation is closely related to the formation and development of diseases. Therefore, how to accurately measure gene variation information within the whole genome has become the key to restricting researchers from exploring the mechanism of disease. WES will design probes for the exon region of the gene, and sequence after capturing the region. Under the condition of constant data throughput, the sequencing depth of the exon region can be significantly improved; however, the disadvantage of this technology is that it cannot monitor non-coding variation located in most regions of the genome, and these non-coding variation often has more significant genetic effect. Studies have shown that WES sequencing data not only contains the sequence of the designed probe site, but also a part of the sequence outside the probe region is captured and then detected by the sequencer. Compared with the probe region, the sequencing depth of the non-probe region / off-target region (Off-target sequence) is often low, and cannot be accurately identified using traditional genome identification algorithms. As large reference genomes become available, new genotype identification algorithms can use reference genotype panels to infer genotypes in regions with low coverage.

[0004] However, large reference genome panels are highly population-specific. Currently, most mainstream databases are based on European or mixed populations. Due to differences in ethnicity leading to differences in haplotypes, the above panels cannot meet the low-coverage genotype identification needs of specific populations in different countries / races / regions. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention provides a non-targeted region genotype filling method and system based on whole exome sequencing; the method of the present invention is based on whole exome sequencing technology and reference whole genome sequencing data set, calculates the genotype possibility of the non-targeted region in the target queue, obtains the genotype estimation result of the non-targeted region, integrates low-depth sequencing data and large reference genome panel, and efficiently and accurately identifies the genotype of genetic variation sites in the non-targeted region of WES; deeply explores the life laws hidden behind the sequencing data and solves related life science problems.

[0006] The first aspect of the present application discloses a non-targeted region genotype filling method based on whole exome sequencing, comprising:

[0007] Obtain whole exome sequencing data of the target cohort and reference whole genome sequencing dataset;

[0008] Filtering the sites in the reference whole genome sequencing data set, and outputting the SNP site information of the reference whole genome sequencing data set; the SNP site information includes chromosome number, genome coordinates and alleles, and population genotype information of sites with non-SNP removed;

[0009] Based on the SNP site information of the reference whole genome sequencing data set and the whole exome sequencing data of the target cohort, the genotype possibility results of each SNP site of each sample in the target cohort are calculated, and the genotype possibility results in each sample are obtained;

[0010] Combining the genotype possibility results in each sample to obtain the genotype possibility results in all samples of the target cohort;

[0011] Performing fragmentation processing on the reference whole genome sequencing data set to obtain genome fragment information after the fragmentation processing;

[0012] Based on the genotype possibility results in all samples of the target cohort, the SNP site information of the reference whole genome sequencing data set, and the genome fragment information, a machine learning algorithm is used to estimate the genotype of the non-targeted region of a single sample in the target cohort to obtain a genotype estimation result of the non-targeted region of a single sample.

[0013] The method further comprises:

[0014] Based on the SNP site information of the reference whole genome sequencing data set and the whole exome sequencing data of the target cohort, the genotype possibility results of each SNP site of each sample in the target cohort are calculated, and the genotype possibility results of the non-targeted region and / or the targeted region in each sample are obtained;

[0015] Combining the genotype possibility results of the non-targeted region and / or the targeted region in each sample to obtain the genotype possibility results of the non-targeted region and / or the targeted region in all samples of the target cohort;

[0016] Performing fragmentation processing on the reference whole genome sequencing data set to obtain genome fragment information after the fragmentation processing;

[0017] Based on the genotype possibility results of the non-targeted regions and / or targeted regions in all samples of the target cohort, the SNP site information of the reference whole genome sequencing data set, and the genome fragment information, a machine learning algorithm is used to estimate the genotype of a single sample in the target cohort, and the genotype of the non-targeted region and / or targeted region of a single sample in the target cohort is estimated to obtain a genotype estimation result of the non-targeted region and / or targeted region of a single sample;

[0018] Optionally, the method for calculating the genotype possibility of each SNP site of each sample in the target queue includes: using mpileup of BCFtools to perform calculations to obtain the genotype possibility of each SNP site calculated based on the sequencing depth;

[0019] Optionally, the estimation results include: imputed genotype dosage, genotype probabilities and best guess genotype results.

[0020] The machine learning algorithm includes but is not limited to the following algorithms to estimate the genotype: iterative optimization algorithm; the iterative optimization algorithm includes one or more of the following: gradient descent, conjugate gradient, coordinate descent, Newton iteration, stepwise regression, minimum angle regression, and Lagrange multiplication.

[0021] The reference whole genome sequencing dataset is a reference whole genome sequencing dataset of a specific population, and the reference whole genome sequencing dataset of the specific population includes specific populations of different countries / races / regions;

[0022] The method further comprises: integrating genome segment information based on the estimation result to obtain chromosome level results including genotype;

[0023] Optionally, the method further comprises: integrating genomic fragment information based on the estimation result to obtain chromosome level results including genotype and haplotype information; the estimation result further comprises: haplotype information.

[0024] Optionally, GLIMPSE_ligated is used to obtain the chromosome level result; the chromosome level result is a VCF file.

[0025] The process of filtering the sites in the reference whole genome sequencing data set and outputting the SNP site information of the reference whole genome sequencing data set includes: extracting bi-allele genetic polymorphism sites (bi-allele SNPs) from the reference whole genome sequencing data set to obtain a first data set and a second data set containing the SNP site information of the reference whole genome sequencing data set;

[0026] Based on the whole exome sequencing data of the target cohort, the first data set and the second data set, the genotype possibility results of each SNP site of each sample in the target cohort are calculated, and the genotype possibility results of the non-targeted region and / or the targeted region in each sample are obtained; the genotype possibility results of the non-targeted region and / or the targeted region in each sample are combined to obtain the genotype possibility results of the non-targeted region and / or the targeted region in all samples of the target cohort;

[0027] Performing fragmentation processing on the reference whole genome sequencing data set to obtain genome fragment information after the fragmentation processing;

[0028] Based on the genotype possibility results of the non-targeted regions and / or targeted regions in all samples of the target cohort, the first data set, and the genomic fragment information, a machine learning algorithm is used to estimate the genotype of a single sample in the target cohort, and the genotype of the non-targeted regions and / or targeted regions of a single sample in the target cohort is estimated to obtain an estimation result of the non-targeted regions and / or targeted regions of a single sample;

[0029] Optionally, the population genotype information from which non-SNP sites have been removed is a first data set; the chromosome number, genome coordinates and alleles are a second data set, and the second data set does not contain genotype information but only contains SNP site information in the first data set.

[0030] The genome segment information includes: the genome segment information after fragmentation processing, including chromosomes, start and end site coordinates.

[0031] The second aspect of the present invention discloses a non-targeted region genotype filling system based on whole exome sequencing, comprising:

[0032] An acquisition unit, used to acquire the whole exome sequencing data of the target cohort and the reference whole genome sequencing data set;

[0033] A first processing unit is used to filter the sites in the reference whole genome sequencing data set, and output the SNP site information of the reference whole genome sequencing data set; the SNP site information includes chromosome number, genome coordinates and alleles, and population genotype information of sites with non-SNP removed;

[0034] The second processing unit calculates the genotype possibility result of each SNP site of each sample in the target cohort based on the SNP site information of the reference whole genome sequencing data set and the whole exome sequencing data of the target cohort, and obtains the genotype possibility result in each sample;

[0035] A third processing unit is used to merge the genotype possibility results in each sample to obtain the genotype possibility results in all samples of the target cohort;

[0036] A fourth processing unit is used to perform fragmentation processing on the reference whole genome sequencing data set to obtain genome fragment information after the fragmentation processing;

[0037] The fifth processing unit estimates the genotype of the non-targeted region of a single sample in the target queue by using a machine learning algorithm based on the genotype possibility results in all samples of the target queue, the SNP site information of the reference whole genome sequencing data set, and the genome fragment information, to obtain the genotype estimation result of the non-targeted region of the single sample.

[0038] The third aspect of the present invention discloses a non-targeted region genotype filling device based on whole exome sequencing, the device comprising: a memory and a processor;

[0039] The memory is used to store program instructions; the processor is used to call the program instructions, and when the program instructions are executed, it is used to execute the above-mentioned non-targeted region genotype filling method based on whole exome sequencing.

[0040] A fourth aspect of the present invention discloses a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the non-targeted region genotype filling method based on whole exome sequencing is implemented.

[0041] Application of the above-mentioned device in diagnosis of disease occurrence and development; optionally, the occurrence and development of the disease is related to changes in the types and quantities of microorganisms; optionally, the disease includes eye diseases (such as myopia), and the WES data of myopia patients are used to fill in non-target areas and / or targeted areas.

[0042] This application has the following beneficial effects:

[0043] 1. The present application innovatively discloses a new non-targeted region genotype filling method, which is based on whole exome sequencing technology and reference whole genome sequencing data sets, calculates the genotype possibility of non-targeted regions and / or targeted regions in the target cohort, obtains genotype estimation results of non-targeted regions and / or targeted regions, integrates low-depth sequencing data and a large reference genome panel (i.e., the reference whole genome sequencing data set in this scheme), and efficiently and accurately identifies the genotypes of genetic variation sites in WES non-targeted regions; deeply explores the life laws hidden behind the sequencing data, ensures the filling of non-targeted regions during subsequent processing, and accurately identifies non-targeted regions. The method overcomes the problem that the existing WES sequencing analysis process cannot identify a large number of genetic variation sites in non-coding regions; at the same time, based on the reference whole genome sequencing data set, it overcomes the problem that the current mainstream database lacks specificity for different national / ethnic / regional populations; when the reference whole genome sequencing data set is a reference whole genome sequencing data set for a population in the same region, the large reference genotype panel of the population in the region is used as a haplotype reference, effectively improving the genotype analysis capability of the regional population cohort; the present invention effectively overcomes the problem that the existing WES sequencing analysis process cannot identify a large number of genetic variation sites in non-coding regions and lacks specificity for the population in the same region, and efficiently and accurately identifies the genotypes of genetic variation sites in non-targeted regions of WES.

[0044] 2. This application innovatively uses an iterative optimization algorithm to estimate the genotype of the sample, combines bioinformatics technology with machine learning technology, and makes full use of existing sequencing data and technical means to solve the current problem of being unable to achieve both cost and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0046] Figure 1 This is a schematic flow chart of a non-targeted region genotype filling method based on whole exome sequencing provided by an embodiment of the present invention;

[0047] Figure 2 Schematic diagram of a non-targeted region genotype filling device based on whole exome sequencing provided by an embodiment of the present invention;

[0048] Figure 3 This is a schematic flow chart of a non-targeted region genotype filling system based on whole exome sequencing provided by an embodiment of the present invention;

[0049] Figure 4This is a flow chart of WES non-targeted region genotype filling provided by an embodiment of the present invention;

[0050] Figure 5 It is a schematic diagram of the distribution of non-targeted region quality control sites on the genome provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0051] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0052] In some of the processes described in the specification and claims of the present invention and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0054] Figure 1 : is a schematic flow chart of a non-targeted region genotype filling method based on whole exome sequencing provided by an embodiment of the present invention. Specifically, the method comprises the following steps:

[0055] 101: Obtain the whole exome sequencing data of the target cohort and the reference whole genome sequencing dataset;

[0056] In one embodiment, the whole exome sequencing data of the target cohort is preprocessed data; the preprocessing process includes: obtaining the whole exome sequencing raw data and human reference genome sequence data of the target cohort (obtaining human reference genome sequence data (fasta format file) from the University of California Genome Research Center (UCSC Genome Browser)); performing quality control on the whole exome sequencing raw data of each sample in the target cohort, and aligning the quality-controlled whole exome sequencing raw data with the human reference genome sequence data (using the genome alignment tool (Burrows-Wheeler Alignment, BWA)) to obtain an alignment result (bam format file - storage of sequencing sequence location information); the alignment result is the whole exome sequencing data of the target cohort. Conventional target region variation detection includes:

[0057] In one embodiment, the reference whole genome sequencing dataset is a reference whole genome sequencing dataset of a specific population, and the reference whole genome sequencing dataset of the specific population includes specific populations of different countries / races / regions; including 24,114,149 genetic polymorphism sites from 12,000 people in the same region, and the whole genome sequencing data of the people in the same region are obtained from a public database, and the genome version number is GRCh37.p5;

[0058] In one embodiment, non-targeted regions in the whole exome sequencing data are identified; identification method: whole exome sequencing technology designs probes for gene exon regions to target and capture the regions, most of which are targeted regions; but some non-targeted sequences will also be captured.

[0059] In one embodiment, the target sample is a sample that needs to be genotyped using this tool, is a cohort sequenced by WES, is the object that the researcher wants to study, and generally includes a disease-health control cohort.

[0060] 102: Filter the sites in the reference whole genome sequencing data set, and output SNP site information of the reference whole genome sequencing data set; the SNP site information includes chromosome number, genome coordinates and alleles, and population genotype information of sites with non-SNP removed;

[0061] In one embodiment, the process of filtering the sites in the reference whole genome sequencing data set and outputting the SNP site information of the reference whole genome sequencing data set includes: extracting bi-allele genetic polymorphism sites (bi-allele SNPs) from the reference whole genome sequencing data set to obtain a first data set and a second data set containing the SNP site information of the reference whole genome sequencing data set. The filtering step is to meet the requirements of calculating GL. The calculation of GL requires all target individuals and all variant sites present in the haplotype reference panel used for interpolation / filling. The tool required to calculate GL is the variation detection file format processing tool-BCFtools. BCFtools cannot accurately calculate the GL of the insertion and deletion of the segment in the genome, so it is necessary to extract the bi-allele genetic polymorphism sites (bi-allele SNPs) in the panel. The variant call format (VCF) of the reference genotype panel is used as input. The file contains a set of genotype information from the same regional population cohort. Use the 'bcftools view' command to extract the bi-allelic sites (SNPs) with the parameters '-m 2 -M 2 -v snps'. After extraction, two output files are obtained, one of which is still in VCF format (the first data set), which is the genetic information of the population without the non-SNP sites. The other output file ends with the suffix 'tsv' (the second data set), does not contain genotype information, and only stores the SNP site information in VCF, including chromosome number, genome coordinates and alleles.

[0062] Based on the whole exome sequencing data of the target cohort, the first data set and the second data set, the genotype possibility results of each SNP site of each sample in the target cohort are calculated, and the genotype possibility results of the non-targeted region and / or the targeted region in each sample are obtained; the genotype possibility results of the non-targeted region and / or the targeted region in each sample are combined to obtain the genotype possibility results of the non-targeted region and / or the targeted region in all samples of the target cohort;

[0063] Performing fragmentation processing on the reference whole genome sequencing data set to obtain genome fragment information after the fragmentation processing;

[0064] Based on the genotype possibility results of the non-targeted regions and / or targeted regions in all samples of the target cohort, the first data set, and the genomic fragment information, a machine learning algorithm is used to estimate the genotype of a single sample in the target cohort, and the genotype of the non-targeted regions and / or targeted regions of a single sample in the target cohort is estimated to obtain an estimation result of the non-targeted regions and / or targeted regions of a single sample;

[0065] In one embodiment, the population genotype information from which non-SNP sites have been removed is a first data set (VCF file); the chromosome number, genomic coordinates and alleles are a second data set (tsv file), and the second data set does not contain genotype information but only contains SNP site information in the first data set.

[0066] In one embodiment, the genome fragment information (text format) includes: the genome fragment information after the fragmentation processing, including chromosomes, start and end site coordinates. During the fragmentation processing, the setting parameters include: the minimum sliding window interval is N (2Mb), and the minimum buffer interval is M (200Kb).

[0067] 103: Based on the SNP site information of the reference whole genome sequencing data set and the whole exome sequencing data of the target cohort, calculate the genotype possibility result of each SNP site of each sample in the target cohort, and obtain the genotype possibility result in each sample;

[0068] In one embodiment, the mpileup tool of BCFtools is used to calculate GL for each sample separately, and the input files are the bam file after genome alignment and the first data set (VCF file) and the second data set (TSV file). The parameters for GL calculation using mpileup of BCFtools are '-I -E -a 'FORMAT / DP''. After calculation, a VCF file is still obtained, which stores the genotype likelihood of each site estimated based on sequencing depth. In one embodiment, genotype likelihood (Genotype Likelihoods, GL) is a measure of the probability that different genotypes may occur, using P-value statistics.

[0069] 104: Combining the genotype possibility results in each sample to obtain the genotype possibility results in all samples of the target cohort;

[0070] In one embodiment, after calculating GL for samples in the target queue in sequence, the merge command of BCFtools is used to merge the GL results of individual samples into an integrated VCF file for subsequent steps.

[0071] 105: performing fragmentation processing on the reference whole genome sequencing data set to obtain genome segmentation information after the fragmentation processing;

[0072] In one embodiment, an important step before constructing haplotypes for genotype interpolation / filling is to define small genomic fragments (chunks). If the region definition is too long, the operation time will be greatly increased, and if it is too short, the accuracy of subsequent genotype identification will be reduced. The GLIMPSE_chunk program of GLIMPSE is used to divide the genome into regions, and the minimum sliding window interval is set to 2Mb and the minimum buffer interval is set to 200Kb. The input file refers to the VCF file (first data set) of the genotype panel, and the output file is in text format. Each line is a divided genome fragment information, including chromosomes, start and end site coordinates.

[0073] 106: Based on the genotype possibility results in all samples of the target cohort, the SNP site information of the reference whole genome sequencing data set, and the genome fragment information, a machine learning algorithm is used to estimate the genotype of the non-targeted region of a single sample in the target cohort to obtain a genotype estimation result of the non-targeted region of the single sample.

[0074] In one embodiment, the genotype likelihood results in all samples of the target cohort are preliminary or non-targeted genotype likelihood results.

[0075] In one embodiment, the estimation results include: imputed genotype dosage, genotype probabilities and best guess genotype results; the same output file contains the above three results.

[0076] In one embodiment, the machine learning algorithm includes but is not limited to the following algorithms to estimate the genotype: iterative optimization algorithm; the iterative optimization algorithm includes one or more of the following: gradient descent, conjugate gradient, coordinate descent, Newton iteration, stepwise regression, minimum angle regression, Lagrange multiplication.

[0077] In one embodiment, the genotype of the non-target region is estimated using the GLIMPSE_phase program, which is optimized for low coverage sequencing data, and the accurate estimation of genotype is performed according to the similarity of the haplotype structure of the reference genotype panel and the haplotype structure of the target queue. The GLIMPSE_phase program is an open source bioinformatics tool, and the iterative optimization algorithm is a step in the internal execution process of the GLIMPSE_phase program; The iterative optimization algorithm is automatically implemented using GLIMPSE_phase. In the present embodiment, the GLIMPSE_phase program is used to realize the estimation of genotype, i.e., directly calling the iterative optimization algorithm to realize the estimation process of genotype; but the genotype estimation is not limited to the GLIMPSE_phase program, i.e., iterative optimization algorithm is not limited to.

[0078] The input data of this step include: the genotype probability results (VCF file) of non-targeted regions and / or targeted regions in all samples of the target cohort, the VCF file of the reference genotype panel (first data set), and the genomic fragment information; the output file format is still VCF, including the best guess genotype, the imputed genotype dosage, and the genotype probabilities. The output result format is 0 / 0:0.038:0.961,0.038,0:0, with colon as the separator. The first information is 0 / 0, which is interpreted as the best guess genotype; the second information is 0.038, which is interpreted as the imputed genotype dosage; the third is 0.961,0.038,0, which is interpreted as the genotype probabilities; the fourth is the haplotype information.

[0079] In one embodiment, the method further comprises:

[0080] Based on the SNP site information of the reference whole genome sequencing data set and the whole exome sequencing data of the target cohort, the genotype possibility results of each SNP site of each sample in the target cohort are calculated, and the genotype possibility results of the non-targeted region and / or the targeted region in each sample are obtained; the genotype possibility results of the non-targeted region and / or the targeted region in each sample are combined to obtain the genotype possibility results of the non-targeted region and / or the targeted region in all samples of the target cohort;

[0081] Performing fragmentation processing on the reference whole genome sequencing data set to obtain genome fragment information after the fragmentation processing;

[0082] Based on the genotype possibility results of the non-targeted regions and / or targeted regions in all samples of the target cohort, the SNP site information of the reference whole genome sequencing data set, and the genomic fragment information, a machine learning algorithm is used to estimate the genotype of a single sample in the target cohort, and the genotype of the non-targeted region and / or targeted region of a single sample in the target cohort is estimated to obtain the genotype estimation results of the non-targeted region and / or targeted region of the single sample.

[0083] In one embodiment, the method further comprises: integrating genomic fragment information based on the estimation result to obtain chromosome level results including genotype;

[0084] Optionally, the method further comprises: integrating genome segment information based on the estimation result to obtain chromosome level results including genotype and haplotype information;

[0085] Optionally, GLIMPSE_ligated is used to obtain the chromosome level result. The chromosome level result is a VCF file. The estimation result also includes: haplotype information. In this case, it is crucial to merge the fragmented genomes together without losing interpolation information. The fragments are integrated using the GLIMPSE_ligated program to obtain a VCF file at the chromosome level, retaining genotype and haplotype information. The input estimation result of this step is in VCF format, and the output is a file after the entire chromosome is connected, which is also in VCF format.

[0086] This application first processes the targeted area and the non-targeted area together, inputs the whole into the calculation GL, and then performs a series of calculations; then the targeted area is processed separately, and the GATK conventional process is used to perform separate SNP calling processing on the targeted area. The present invention performs genotype filling on the non-targeted area of ​​the target queue. The present invention performs genotype filling on the non-targeted area of ​​the target queue. As a tool set for diversity sites, the typical process of GATK mainly consists of 7 steps: the first part is sequence alignment, which is mainly completed by BWA; the second part is data cleaning, which mainly includes the steps of MarkDuplicates, Sort, Indel Realignment and Base Recalibration; the third part is diversity discovery, which is mainly composed of HaplotypeCaller and the subsequent Joint Genotyping and VariantRecalibration.

[0087] Genes are composed of thousands of nucleotide pairs. The nucleotide sequences that make up genes can be divided into different segments. In the process of gene expression, different segments play different roles. The segment that can be transcribed into the corresponding messenger RNA and then guide protein synthesis (that is, it can encode proteins) is called the coding region. The segment that cannot encode proteins is called the non-coding region. The non-coding region is located before and after the coding region, belongs to the same gene, and controls the expression and strength of the gene.

[0088] Although non-coding regions cannot encode proteins, they are indispensable for the expression of genetic information. They contain nucleotide sequences that regulate the expression of genetic information and have genetic effects, such as RNA polymerase binding sites. Mutations in non-coding regions will not affect genetics.

[0089] The non-coding region is indispensable to the target gene. There are binding sites for RNA polymerase on the non-coding region, which has a regulatory effect.

[0090] Insertions, deletions, and substitutions of bases in the non-coding region of a gene are also considered gene mutation events, although most studies are limited to mutations in the coding region. Promoters are located in non-coding regions. Because of the special arrangement of palindromic sequences, most of them are located in non-coding regions.

[0091] Figure 2 A non-targeted region genotype filling device based on whole exome sequencing is provided in an embodiment of the present invention, and the device comprises: a memory and a processor; the memory is used to store program instructions; the processor is used to call the program instructions, and when the program instructions are executed, it is used to execute the above-mentioned non-targeted region genotype filling method based on whole exome sequencing.

[0092] Figure 3 A non-targeted region genotype filling system based on whole exome sequencing is provided in an embodiment of the present invention, comprising:

[0093] An acquisition unit 301 is used to acquire the whole exome sequencing data of the target cohort and the reference whole genome sequencing data set;

[0094] The first processing unit 302 is used to filter the sites in the reference whole genome sequencing data set and output the SNP site information of the reference whole genome sequencing data set; the SNP site information includes chromosome number, genome coordinates and alleles, and population genotype information of sites with non-SNP removed;

[0095] The second processing unit 303 calculates the genotype possibility result of each SNP site of each sample in the target cohort based on the SNP site information of the reference whole genome sequencing data set and the whole exome sequencing data of the target cohort, and obtains the genotype possibility result in each sample;

[0096] The third processing unit 304 is used to merge the genotype possibility results in each sample to obtain the genotype possibility results in all samples of the target queue;

[0097] The fourth processing unit 305 is used to perform fragmentation processing on the reference whole genome sequencing data set to obtain genome fragment information after the fragmentation processing;

[0098] The fifth processing unit 306 estimates the genotype of the non-targeted region of a single sample in the target queue using a machine learning algorithm based on the genotype possibility results in all samples in the target queue, the SNP site information of the reference whole genome sequencing data set, and the genome fragment information to obtain the genotype estimation result of the non-targeted region of the single sample.

[0099] The embodiment of the present invention also provides a non-targeted region genotype filling system for whole exome sequencing, comprising:

[0100] A first acquisition unit is used to acquire a reference whole genome sequencing data set;

[0101] A first processing unit is used to extract bi-allele genetic polymorphism sites (bi-allele SNPs) from the reference whole genome sequencing data set to obtain a first data set and a second data set;

[0102] A second acquisition unit is used to acquire whole exome sequencing data of a target cohort;

[0103] The second processing unit is used to filter the sites in the reference whole genome sequencing data set, and output the SNP site information of the reference whole genome sequencing data set; the SNP site information includes chromosome number, genome coordinates and alleles, and population genotype information of sites without SNPs; the population genotype information of sites without SNPs is the first data set (VCF file); the chromosome number, genome coordinates and alleles are the second data set (tsv file), and the second data set does not include genotype information, but only includes the SNP site information in the first data set;

[0104] A third processing unit calculates the genotype possibility results of each SNP site of each sample in the target cohort based on the whole exome sequencing data of the target cohort, the first data set and the second data set, and obtains the genotype possibility results of the non-targeted region and / or the targeted region in each sample;

[0105] A fourth processing unit is used to merge the genotype possibility results of the non-targeted region and / or the targeted region in each sample to obtain the genotype possibility results of the non-targeted region and / or the targeted region in all samples of the target cohort;

[0106] A fifth processing unit is used to perform fragmentation processing on the reference whole genome sequencing data set to obtain genome fragment information after the fragmentation processing;

[0107] The sixth processing unit estimates the genotype of the non-targeted area and / or the targeted area of ​​a single sample in the target queue by using a machine learning algorithm based on the genotype possibility results of the non-targeted area and / or the targeted area in all samples of the target queue, the first data set and the genomic fragment information, to obtain the estimation result of the non-targeted area and / or the targeted area of ​​the single sample.

[0108] Figure 4 It is a flowchart of filling in WES non-targeted region genotypes provided in an embodiment of the present invention, which mainly involves a process of preprocessing the whole exome sequencing data of the target cohort (library construction, comparison with the reference genome, etc.), and then calculating the GL in the targeted region and / or non-targeted region of each sample in the target cohort based on the large reference genome panel, and then merging them to finally obtain the genotype estimation results of the targeted region and / or non-targeted region of a single sample.

[0109] Experimental results:

[0110] (1) Statistics of sequencing coverage of non-targeted regions by WES;

[0111] a) Statistics of the coverage of standard WES sequencing data showed that the average coverage of the exon probe region was 77, and the average coverage of the non-probe region was 0.87;

[0112] b) The non-probe regions had coverages of 5%, 3.75%, and 3.19% at 1X, 3X, and 5X, respectively, meaning that the number of sites with a coverage above 3X was 4,910,578.

[0113] (2) Frequency distribution of regional populations and loci with coverage above 3X in the 1000 Genomes panel;

[0114] a) According to the statistics of regional population genotype panel, the number of sites with sequencing depth of more than 3X is 4,910,578, and according to the statistics of 1000 Genomes panel, the number of sites with sequencing depth of more than 3X is 16,919,510, and the number of sites shared by the two is 3,419,584. Since the population abundance in 1000 Genomes is higher, the number of polymorphic sites is more, but there are still some sites that are specific to the population in the same region.

[0115] (3) Correlation between genotype likelihood (GL) quality value and sequence coverage;

[0116] When using the genotype panel of the same regional population, the genotype likelihood (GL) calculation quality value is directly correlated with the sequence coverage (correlation coefficient is 0.676)

[0117] (4) Distribution of quality values ​​(INFO score) after genotype imputation / filling in non-targeted regions;

[0118] According to the reference genotype panel (whole genome sequencing data set) of the same regional population, a total of 23,639,100 polymorphic sites were obtained in the non-targeted region after interpolation, and the quality value distribution is shown in the following table:

[0119]

[0120] Due to the high sequencing coverage, the INFO value of the WES targeted region is high, but 66.91% of the sites in the non-targeted region still have INFO values ​​above 0.9. If the post-interpolation quality value INFO>0.9 and sequencing coverage>3X are used as screening conditions, 4,910,579 polymorphic sites can be obtained.

[0121] (5) Distribution of sites on the genome after quality control of non-targeted regions;

[0122] After analyzing the genomic distribution of 4,910,579 polymorphic sites after quality control, it was found that nearly half of the sites were still located in the protein coding region (48.56%), followed by the UTR (untranslated region) and promoter region, and about 8.49% of the sites were located in the enhancer region. Figure 5 As shown, the distribution of polymorphic sites in CDS, Enhancer, LncRNA, Promoter, Others, 3UTR, and 5UTR are shown respectively.

[0123] The present invention effectively utilizes the low coverage site information of the non-targeted region of whole exome sequencing, combined with the reference genotype panel of the same regional population, can significantly increase the number of polymorphic sites found by WES sequencing, cover more genomic intervals, and provide more data for genetic association analysis. After analysis, if strict sequencing coverage>3X and INFO>0.9 are used as thresholds, the present invention can expand the number of genotype sites of traditional WES sequencing data by 4910579, saving costs and obtaining greater benefits.

[0124] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned non-targeted region genotype filling method based on whole exome sequencing.

[0125] The validation results of this validation example show that assigning intrinsic weights to indications can moderately improve the performance of the method relative to the default setting.

[0126] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0127] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0128] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0129] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0130] A person skilled in the art may understand that all or part of the steps in the various methods of the above embodiments may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the storage medium may include: a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc.

[0131] A person skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. The above-mentioned storage medium can be a read-only memory, a disk or an optical disk, etc.

[0132] The above is a detailed introduction to a computer device provided by the present invention. For a person skilled in the art, according to the concept of the embodiments of the present invention, there may be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. Non-targeted region genotype filling method based on whole exome sequencing, including: Obtain whole exome sequencing data of the target cohort and reference whole genome sequencing dataset; Filtering the sites in the reference whole genome sequencing dataset, and outputting SNP site information of the reference whole genome sequencing dataset; The SNP site information includes chromosome number, genome coordinates and alleles, and population genotype information of sites with non-SNP removed; Based on the SNP site information of the reference whole genome sequencing data set and the whole exome sequencing data of the target cohort, the genotype possibility results of each SNP site of each sample in the target cohort are calculated, and the genotype possibility results in each sample are obtained; Combining the genotype possibility results in each sample to obtain the genotype possibility results in all samples of the target cohort; Performing fragmentation processing on the reference whole genome sequencing data set to obtain genome fragment information after the fragmentation processing; Based on the genotype possibility results in all samples of the target cohort, the SNP site information of the reference whole genome sequencing data set, and the genome fragment information, a machine learning algorithm is used to estimate the genotype of the non-targeted region of a single sample in the target cohort to obtain a genotype estimation result of the non-targeted region of a single sample.

2. The non-targeted region genotype filling method based on whole exome sequencing according to claim 1, characterized in that: The method further comprises: Based on the SNP site information of the reference whole genome sequencing data set and the whole exome sequencing data of the target cohort, the genotype possibility results of each SNP site of each sample in the target cohort are calculated, and the genotype possibility results of the non-targeted region and / or the targeted region in each sample are obtained; Combining the genotype possibility results of the non-targeted region and / or the targeted region in each sample to obtain the genotype possibility results of the non-targeted region and / or the targeted region in all samples of the target cohort; Performing fragmentation processing on the reference whole genome sequencing data set to obtain genome fragment information after the fragmentation processing; Based on the genotype possibility results of the non-targeted regions and / or targeted regions in all samples of the target cohort, the SNP site information of the reference whole genome sequencing data set, and the genomic fragment information, a machine learning algorithm is used to estimate the genotype of a single sample in the target cohort, and the genotype of the non-targeted region and / or targeted region of a single sample in the target cohort is estimated to obtain the genotype estimation results of the non-targeted region and / or targeted region of the single sample.

3. The non-targeted region genotype filling method based on whole exome sequencing according to claim 2, characterized in that: The method for calculating the genotype possibility of each SNP site of each sample in the target queue includes: using mpileup of BCFtools to perform calculations to obtain the genotype possibility of each SNP site calculated based on sequencing depth.

4. The non-targeted region genotype filling method based on whole exome sequencing according to claim 2, characterized in that: The estimation results include: genotype dosage after filling, genotype posterior probability and the result of the best estimated genotype.

5. The method for filling non-targeted region genotypes based on whole exome sequencing according to claim 1, characterized in that: The machine learning algorithm includes the following algorithms to estimate the genotype: iterative optimization algorithm; the iterative optimization algorithm includes one or more of the following: gradient descent, conjugate gradient, coordinate descent, Newton iteration, stepwise regression, minimum angle regression, and Lagrange multiplication.

6. The method for filling non-targeted region genotypes based on whole exome sequencing according to claim 1, characterized in that: The reference whole genome sequencing dataset is a reference whole genome sequencing dataset of a specific population, and the reference whole genome sequencing dataset of the specific population includes specific populations of different countries / races / regions.

7. The method for filling non-targeted region genotypes based on whole exome sequencing according to claim 1, characterized in that: The method further comprises: integrating genome fragment information based on the estimation result to obtain chromosome level results including genotype.

8. The method for filling non-targeted region genotypes based on whole exome sequencing according to claim 1, characterized in that: The method further comprises: integrating genome segment information based on the estimation result to obtain chromosome level results including genotype and haplotype information.

9. The method for filling non-targeted region genotypes based on whole exome sequencing according to claim 8, characterized in that: The chromosome level results were obtained using GLIMPSE_ligated.

10. The method for filling non-targeted region genotypes based on whole exome sequencing according to claim 1, characterized in that: The process of filtering the sites in the reference whole genome sequencing data set and outputting the SNP site information of the reference whole genome sequencing data set comprises: extracting biallelic genetic polymorphism sites from the reference whole genome sequencing data set to obtain a first data set and a second data set containing the SNP site information of the reference whole genome sequencing data set; Based on the whole exome sequencing data of the target cohort, the first data set and the second data set, the genotype possibility results of each SNP site of each sample in the target cohort are calculated, and the genotype possibility results of the non-targeted region and / or the targeted region in each sample are obtained; the genotype possibility results of the non-targeted region and / or the targeted region in each sample are combined to obtain the genotype possibility results of the non-targeted region and / or the targeted region in all samples of the target cohort; Performing fragmentation processing on the reference whole genome sequencing data set to obtain genome fragment information after the fragmentation processing; Based on the genotype possibility results of the non-targeted regions and / or targeted regions in all samples of the target cohort, the first data set and the genomic fragment information, a machine learning algorithm is used to estimate the genotype of a single sample in the target cohort, and the genotype of the non-targeted regions and / or targeted regions of a single sample in the target cohort is estimated to obtain an estimation result of the non-targeted regions and / or targeted regions of a single sample.

11. The method for filling non-targeted region genotypes based on whole exome sequencing according to claim 10, characterized in that: The population genotype information from which non-SNP sites have been removed is the first data set; the chromosome number, genome coordinates and alleles are the second data set, which does not include genotype information but only includes the SNP site information in the first data set.

12. The method for filling non-targeted region genotypes based on whole exome sequencing according to claim 1, characterized in that: The genome segment information includes: the genome segment information after fragmentation processing, including chromosomes, start and end site coordinates.

13. A non-targeted region genotype filling system based on whole exome sequencing, comprising: An acquisition unit, used to acquire the whole exome sequencing data of the target cohort and the reference whole genome sequencing data set; A first processing unit is used to filter the sites in the reference whole genome sequencing data set and output SNP site information of the reference whole genome sequencing data set; The SNP site information includes chromosome number, genome coordinates and alleles, and population genotype information of sites with non-SNP removed; The second processing unit calculates the genotype possibility result of each SNP site of each sample in the target cohort based on the SNP site information of the reference whole genome sequencing data set and the whole exome sequencing data of the target cohort, and obtains the genotype possibility result in each sample; A third processing unit is used to merge the genotype possibility results in each sample to obtain the genotype possibility results in all samples of the target cohort; A fourth processing unit is used to perform fragmentation processing on the reference whole genome sequencing data set to obtain genome fragment information after the fragmentation processing; The fifth processing unit estimates the genotype of the non-targeted region of a single sample in the target queue by using a machine learning algorithm based on the genotype possibility results in all samples of the target queue, the SNP site information of the reference whole genome sequencing data set, and the genome fragment information, to obtain the genotype estimation result of the non-targeted region of the single sample.

14. A non-targeted region genotype filling device based on whole exome sequencing, the device comprising: Memory and processor; The memory is used to store program instructions; The processor is used to call program instructions, and when the program instructions are executed, it is used to execute the non-targeted region genotype filling method based on whole exome sequencing according to any one of claims 1-12.

15. A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the non-targeted region genotype filling method based on whole exome sequencing according to any one of claims 1 to 12 is implemented.

Images