A hybrid DNA probabilistic genotyping method for MH sequencing data

Abstract

The present invention discloses a hybrid DNA probability typing method for MH sequencing data. The method comprises the following steps: step 1: obtaining a priori parameters according to MH sequencing data of DNA from a single individual; step 2: obtaining characteristic information of MH sequencing data of a mixed DNA sample to be tested; step 3: filtering out noise sequences according to the sequencing information; step 4: constructing likelihood functions respectively under the prosecution hypothesis and the defense hypothesis according to the priori parameters and characteristic information, and calculating the likelihood ratio of the prosecution hypothesis and the defense hypothesis; step 5: sorting possible genotype combinations according to the value of the likelihood function; the present invention is used for the detection of common mixed DNA samples in forensic DNA analysis work, and can objectively and accurately analyze the results of MH sequencing data of mixed DNA, effectively reducing the uncertainty and subjective errors in the analysis process, and significantly improving the accuracy and scientificity of MH-MPS data analysis of forensic mixed DNA.

Description

Technical Field

[0001] The present invention relates to the technical field of forensic DNA analysis, and in particular to a hybrid DNA probabilistic typing method for MH sequencing data. Background Art

[0002] Mixed DNA refers to a DNA sample that is a mixture of DNA from multiple individuals. It is a common and most difficult sample type to analyze in forensic DNA analysis. Each individual that makes up the mixed DNA is also called a contributor. The genotyping results of the genetic markers of the mixed DNA are expressed as mixed genotypes, that is, the alleles of multiple individuals are mixed together. The main purposes of mixed DNA analysis are: 1) to separate the possible individual genotypes of each contributor from the mixed genotyping results, a process that can be called "deconvolution". 2) to use biostatistical methods to evaluate the mixed genotyping results as forensic science evidence and the ability to prove the facts of the case, that is, "evidence strength evaluation".

[0003] Microhaplotype (MH) is a new type of genetic marker based on DNA sequence polymorphism that has attracted much attention in the field of forensic medicine. It is usually detected on a massive parallel sequencing (MPS) platform to achieve genotyping. MH has the characteristics of good polymorphism, high sensitivity, low mutation rate, short fragment length, and no shadow peak products. These characteristics make MH suitable for forensic analysis of complex biological samples, especially genotyping of mixed DNA, and provide a new solution for forensic DNA analysis of mixed biological samples of multiple individuals.

[0004] In recent years, more and more studies have focused on the use of MH genetic markers to genotype forensic mixed DNA samples. These studies not only confirm the application potential of MH in the field of forensics, but also provide a theoretical basis for its application in actual cases. However, due to the complexity of the second-generation sequencing technology process, as well as the complexity of multi-site complex targeted amplification and sequencing data structure of MH genetic markers, there is currently a lack of analysis methods for MH second-generation sequencing data of mixed DNA samples.

[0005] Existing MH analysis methods mainly analyze MH sequencing results through observation and subjective judgment of analysts; therefore, the results of mixed MH sequencing data analysis are mainly qualitative, with poor reliability and accuracy. Summary of the invention

[0006] In view of the problems existing in the prior art, the present invention proposes a hybrid DNA probabilistic typing method for MH sequencing data.

[0007] The technical solution adopted by the present invention is: a hybrid DNA probabilistic typing method for MH sequencing data, comprising the following steps:

[0008] Step 1: construct a training set based on the MH sequencing data of a single individual's DNA, and estimate the prior parameters based on the training set;

[0009] Step 2: Obtain characteristic information of MH sequencing data of the mixed DNA sample to be tested, including MH typing and quantitative information of the mixed DNA, typing information of the target individual, typing information of the reference individual and related parameters;

[0010] Step 3: Filter out noise sequences based on sequencing information;

[0011] Step 4: Based on the prior parameters obtained in step 1 and the feature information obtained in step 2, construct likelihood functions under the prosecution hypothesis and the defense hypothesis respectively, and calculate the likelihood ratio of the prosecution hypothesis and the defense hypothesis based on the likelihood function;

[0012] The prosecution hypothesis is: the target individual and n unknown individuals are contributors to the mixed DNA sample, or the target individual, the reference individual, and n-1 unknown individuals are contributors to the mixed DNA sample;

[0013] The defense hypothesis is: n+1 unknown individuals are contributors to the mixed DNA sample, or the reference individual and n unknown individuals are contributors to the mixed sample;

[0014] The target individual is the individual suspected to be one of the contributors to the mixed DNA sample;

[0015] A reference individual is an individual that has been identified as one of the contributors to a mixed DNA sample;

[0016] Step 5: Calculate the likelihood function values ​​of all genotype combinations at each locus and obtain the ranking results of the locus likelihood function values.

[0017] Furthermore, the a priori parameters in step 1 include: locus-specific detection efficacy parameters of MH genetic markers, analysis threshold, allele dropout rate, allele population frequency, and noise distribution parameters.

[0018] Further, the MH sequencing data of the mixed DNA in step 2 includes: MH locus, total sequencing reads of MH locus, MH sequence base information, and MH sequence sequencing reads;

[0019] The target individual typing information includes: the MH allele base information of the individual at each MH locus;

[0020] The reference individual typing information includes: the MH allele base information of the individual at each MH locus;

[0021] Relevant parameters include: prior parameters and number of contributors.

[0022] Furthermore, the process of filtering out the noise sequence in step 3 is as follows:

[0023] According to the MH sequencing data of the mixed DNA sample to be tested, the signal ratio r of each base sequence is calculated:

[0024] The MH sequencing data of the mixed DNA sample to be tested is filtered according to the analysis threshold, and those that are not within the threshold range are filtered out.

[0025] Furthermore, the likelihood function construction method in step 4 is as follows:

[0026] S1: Obtain all possible genotype combinations at each locus;

[0027] S2: Calculate the genotype frequency of each contributor based on the allele population frequency information;

[0028] S3: Obtain the corresponding weights based on the genotype combination and the number of MH sequence sequencing reads;

[0029] S4: Calculate the likelihood function Lik on a locus l l :

[0030]

[0031] Where: i is the genotype combination number, G i is the ith genotype combination, P(G i |H) is G i The corresponding frequency in the group, H is the prosecution hypothesis or the defense hypothesis, D Gil is the ith genotype combination G i The weight of

[0032] S5: Calculate the likelihood function Lik of all loci of the sample to be tested:

[0033]

[0034] Where l is the locus number and L is the number of all loci.

[0035] Furthermore, the process of calculating the genotype frequency of each contributor in step S2 is as follows:

[0036] The genotype frequency of each contributor was calculated based on the allele population frequency information and the Hardy-Weinberg equilibrium law.

[0037] Furthermore, the process of obtaining the corresponding weight according to the genotype combination and the MH sequence sequencing reading in step S3 is as follows:

[0038] Get the sequencing read information O of the set O of all sequences at the specified locus l c Sum signal ratio O r ;

[0039] In the specified genotype combination G i Calculate the effective number of alleles A for each sequence in set O, where the effective number of alleles of sequence o is A o ;

[0040]

[0041] Where: B Ox is the number of sequences o in the sequence set O contained in the genotype of the xth contributor, M x is the mixing ratio parameter of the x-th contributor, x is the contributor number, and X is the number of contributors;

[0042] According to the significant number of alleles A o Classify all sequences. If A o If it is not 0, it is the allele sequence. o If it is 0, it is a noise sequence, and the sequence included in the genotype combination but not included in all sequence sets O is a missing allele sequence;

[0043] The weights of the allele sequence, noise sequence, and missing allele sequence are calculated separately, and then the weights are multiplied to obtain the weight corresponding to the genotype combination.

[0044] Furthermore, the process of calculating the likelihood ratio of the prosecution hypothesis and the defense hypothesis according to the likelihood function in step 4 is as follows:

[0045] Parameters of the likelihood function under the prosecution hypothesis and the defense hypothesis are estimated respectively;

[0046] According to the estimated parameter values, the size of the likelihood function is calculated under the prosecution hypothesis and the defense hypothesis respectively, and the maximum value of the likelihood function under the prosecution hypothesis and the defense hypothesis is taken respectively;

[0047] Get the likelihood ratio between the prosecution hypothesis and the defense hypothesis.

[0048] Furthermore, the estimated parameters include: a normal distribution homogeneity parameter, a normal distribution variability parameter and a mixing ratio parameter of each contributor.

[0049] Furthermore, the weight D of the missing allele sequence o for:

[0050] D o =(P(Dropout)) z

[0051] Where: P (Dropout) is the allele dropout rate, z is the number of allele dropouts in the genotype combination;

[0052] The weight of the noise sequence D no for:

[0053]

[0054] Where: k is the noise sequence number, K is the number of noise sequences, r k is the signal ratio of the kth noise sequence at the locus, τ1 and τ2 are noise distribution parameters, and beta represents the probability density function of the beta distribution;

[0055] Weight of allele sequence D a for:

[0056]

[0057] Sequencing reads y for the nth allele at a locus n Normal distribution

[0058] y n :norm(A n *α l μ,CV*A n *α l μ);

[0059] Where: α l is the specific detection efficiency parameter of locus l, μ is the mean parameter of normal distribution, CV is the variability parameter of normal distribution, n is the allele sequence number, N is the number of allele sequences at the locus, and A n is the significant number of alleles of the nth allele, and norm represents the probability density function of the normal distribution.

[0060] The beneficial effects of the present invention are:

[0061] (1) The analysis method of the present invention more objectively and accurately analyzes the results of MH sequencing data of mixed DNA, effectively reduces the uncertainty and subjective errors in the analysis process, and significantly improves the accuracy and scientificity of MH analysis of mixed DNA;

[0062] (2) The present invention improves the efficiency of analysis and can realize the automation of analysis, making complex data processing more efficient without the need for manual analysis;

[0063] (3) The present invention can analyze more sample information in the analysis of mixed and complex mixed MH sequencing data; it can deeply mine and interpret more information hidden in the sequencing data;

[0064] (4) The analytical method of the present invention promotes the practical application of MH genetic markers in the field of forensic DNA analysis and lays a foundation for its future development. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0066] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0067] like Figure 1 As shown, a hybrid DNA probabilistic typing method for MH sequencing data comprises the following steps:

[0068] Step 1: construct a training set based on the MH sequencing data of a single individual's DNA, and estimate the prior parameters based on the training set;

[0069] The training set data is a set of MH sequencing data of DNA from a single individual detected under the same conditions, including: MH detection reagents, MPS library preparation process, MPS sequencing program, sequencing data analysis process, etc. The training set data is used to estimate a priori parameters, including: locus-specific detection efficiency parameters of MH genetic markers, analysis thresholds, allele dropout rates, allele population frequencies, and sequencing noise distribution parameters.

[0070] The estimated prior parameter values ​​can be used for the analysis of all unknown samples tested under the same conditions. Without significantly changing the test conditions, there is no need to re-estimate the prior parameters for each analysis.

[0071] Step 2: Obtain characteristic information of the MH sequencing data of the sample to be tested, including MH typing and quantitative information of the mixed DNA, typing information of the target individual, typing information of the reference individual, and related parameters;

[0072] The target individual is the individual suspected to be one of the contributors to the mixed DNA sample;

[0073] A reference individual is an individual that has been identified as one of the contributors to a mixed DNA sample.

[0074] MH sequencing data of mixed DNA include: MH locus, total sequencing reads of MH locus, MH sequence base information, and MH sequence sequencing reads;

[0075] The target individual typing information includes: the MH allele base information of the individual at each MH locus;

[0076] The reference individual typing information includes: the MH allele base information of the individual at each MH locus;

[0077] Relevant parameters include: prior parameters and number of contributors.

[0078] Step 3: Filter out noise sequences based on sequencing information;

[0079] The process of filtering out noise sequence is as follows:

[0080] According to the MH sequencing data of the mixed DNA sample to be tested, the signal ratio r of each base sequence is calculated;

[0081]

[0082] Where: d is the sequence sequencing reads, and D is the total sequencing reads of the locus.

[0083] The signal ratio of each base sequence in the MH sequencing data of the mixed DNA sample to be tested is filtered according to the analysis threshold, and those that are not within the threshold range are filtered out, and only the base sequences with r greater than the analysis threshold are retained.

[0084] Step 4: Based on the prior parameters obtained in step 1 and the characteristic information obtained in step 2, construct likelihood functions under the prosecution hypothesis and the defense hypothesis respectively, and calculate the likelihood ratio of the prosecution hypothesis and the defense hypothesis based on the likelihood function; the likelihood ratio is used as the strength of evidence of the mixed DNA sample to be tested.

[0085] The prosecution hypothesis is: the target individual and n unknown individuals are contributors to the mixed DNA sample, or the target individual, the reference individual, and n-1 unknown individuals are contributors to the mixed DNA sample;

[0086] The defense hypothesis is: n+1 unknown individuals are contributors to the mixed DNA sample, or the reference individual and n unknown individuals are contributors to the mixed sample;

[0087] The likelihood function construction process is as follows:

[0088] S1: Obtain all possible genotype combinations at each locus;

[0089] According to the prosecution hypothesis and the defense hypothesis, the MH sequencing data of the mixed DNA sample to be tested, and the typing information of the target individual and the reference individual, all possible genotype combinations G are listed for each locus under each hypothesis.

[0090] First, extract all unfiltered base sequences O on the specified locus l, and form a candidate allele set a = {O, Q} with an allele dropout placeholder Q. For each unknown contributor, extract two candidate alleles from set a with replacement to form its genotype. All possible genotypes of the unknown contributors are combined in the form of a Cartesian product with replacement. If there is a target individual or reference individual under this hypothesis, the genotype combination of the unknown individual that has been constructed is combined with the known genotype of the target individual or reference individual at the locus to form all possible genotype combinations G at the locus.

[0091] S2: Calculate the genotype frequency of each contributor based on the allele population frequency information;

[0092] The genotype frequency of each contributor is calculated based on the allele population frequency information and the Hardy-Weinberg equilibrium law: When the contributor is heterozygous, such as AB, its genotype frequency is 2p A p B , where p A and p B are the frequencies of alleles A and B in the population, respectively. When the contributor is homozygous, such as AA, its genotype frequency is 2p A 2 . Then the frequency P(G i |H) is the product of the genotype frequencies of all contributors at that locus.

[0093] S3: Obtain the corresponding weights based on the genotype combination and the number of MH sequence sequencing reads;

[0094] The corresponding probability or density is calculated based on all possible genotype combinations at each locus and the sequencing read information of the MH allele sequence, which is called weight.

[0095] First, the sequencing read information O of all sequences O on the specified locus l is obtained c Sum signal ratio O r .

[0096] In the specified genotype combination G i Calculate the effective number of alleles A for each sequence O, where the effective number of alleles of sequence o is A o

[0097]

[0098] Where: B Ox is the number of sequences o in the sequence set O contained in the genotype of the xth contributor, M xis the mixing ratio parameter of the x-th contributor, x is the contributor number, and X is the number of contributors; compare each sequence o in the set O with the genotype of each contributor in the genotype combination in turn to obtain the contribution of each contributor to o, and then add up the contributions of all contributors to obtain the effective number of alleles of sequence o.

[0099] According to the significant number of alleles A o Classify all sequences O, if A o If it is not 0, it is the allele sequence. o If it is 0, it is a noise sequence, and the sequence included in the genotype combination but not included in all sequence sets O is a missing allele sequence;

[0100] The weights of the allele sequence, noise sequence, and missing allele sequence are calculated separately, and then the weights are multiplied to obtain the weight corresponding to the genotype combination.

[0101] The weight of the missing allele sequence D o for:

[0102] D o =(P(Dropout)) z

[0103] Where: P (Dropout) is the allele dropout rate, z is the number of allele dropouts in the genotype combination, the loss of one allele in a heterozygote is counted as 1 time, and the loss of a homozygote is counted as 2 times.

[0104] The weight of the noise sequence D no is the product of the Gehry density functions of each noise sequence:

[0105]

[0106] Where: k is the noise sequence number, K is the number of noise sequences, r k is the signal ratio of the kth noise sequence at the locus, r k :beta(τ1,τ2), τ1 and τ2 are noise distribution parameters, beta represents the probability density function of beta distribution;

[0107] Weight of allele sequence D a is the product of the probability density functions for each allele sequence:

[0108]

[0109] Sequencing reads y for the nth allele at a locus n Normal distribution

[0110] y n :norm(An *α l μ,CV*A n *α l μ);

[0111] Where: α l is the specific detection efficiency parameter of locus l, μ is the mean parameter of normal distribution, CV is the variability parameter of normal distribution, n is the allele sequence number, N is the number of allele sequences at the locus, and A n is the significant number of alleles of the nth allele, and norm represents the probability density function of the normal distribution.

[0112] Then at the specified locus l, the i-th genotype combination G i Weight calculation for:

[0113]

[0114] S4: Calculate the likelihood function Lik on a locus l l :

[0115]

[0116] Where: i is the genotype combination number, G i is the ith genotype combination, P(G i |H) is the frequency of Gi in the group, H is the prosecution hypothesis or the defense hypothesis, is the ith genotype combination G i The weight of

[0117] S5: The likelihood function Lik of all loci of the sample to be tested under the prosecution hypothesis and the defense hypothesis:

[0118]

[0119] Where l is the locus number and L is the number of all loci.

[0120] The process of calculating the likelihood ratio of the prosecution hypothesis and the defense hypothesis based on the likelihood function is as follows:

[0121] Parameters of the likelihood function under the prosecution hypothesis and the defense hypothesis are estimated respectively; the estimated parameters include: normal distribution homogeneity parameter μ, normal distribution variability parameter CV and mixing ratio parameter M of each contributor x , the number of mixing ratio parameters is the number of assumed contributors minus 1. The maximum likelihood estimation method is used to estimate the value of the parameter, and each sample is estimated three times.

[0122] According to the estimated parameter values ​​(three sets of parameter values), the size of the likelihood function is calculated under the prosecution hypothesis and the defense hypothesis respectively, and the maximum value of the likelihood function under the prosecution hypothesis and the defense hypothesis is taken respectively;

[0123] Get the likelihood ratio between the prosecution hypothesis and the defense hypothesis.

[0124] Step 5: Calculate the likelihood function values ​​of all genotype combinations at each locus and obtain the result of the size ranking of the locus likelihood function values. This result provides a basis for the deconvolution of the mixed DNA to be tested, thereby achieving the purpose of forensic mixed DNA analysis and interpretation.

[0125] The present invention will be further described below through specific data.

[0126] Step 1: Use MH sequencing data of 100 single-source DNA samples with a sample size of 1 ng as the training set to estimate the value of the prior parameter. The training set comes from 100 different individuals, and genotyping is performed under the same MH detection reagents, MPS library preparation process, MPS sequencing program, and sequencing data analysis process.

[0127] The obtained prior parameters are as follows:

[0128] The allele dropout rate is 0.05, the noise distribution parameters τ1 = 0.32, τ2 = 530.19; the value of the locus-specific detection efficiency parameter α l As shown in Table 1; the analysis threshold was 0.02, and the population frequencies of the MH alleles used were shown in Table 2.

[0129] Table 1. Locus-specific assay performance parameters

[0130]

[0131]

[0132] Table 2. Population frequency of MH alleles

[0133]

[0134]

[0135]

[0136]

[0137]

[0138] Step 2: Obtain characteristic information of MH sequencing data of the mixed DNA sample to be tested, including MH typing and quantitative information of the mixed DNA, typing information of the target individual, typing information of the reference individual and related parameters;

[0139] The MH typing and quantitative information of mixed DNA includes MH loci, total sequencing reads of MH loci, MH sequence base information, and MH sequence sequencing reads.

[0140] The typing information of the target individual includes the MH allele base information of the individual at each MH locus.

[0141] The reference individual typing information includes: the MH allele base information of the individual at each MH locus;

[0142] Related parameters, including the values ​​of the aforementioned prior parameters, and the number of contributors is set to 2.

[0143] Step 3: Filter out noise sequences based on sequencing information

[0144] According to the MH sequencing data of mixed DNA, the signal ratio r of each base sequence was calculated;

[0145] The signal ratio r of each base sequence in the MH sequencing data of mixed DNA was filtered according to the analysis threshold of 0.01, and only the base sequences with r greater than 0.01 were retained.

[0146] Step 4: Based on the prior parameters obtained in step 1 and the feature information obtained in step 2, construct likelihood functions under the prosecution hypothesis and the defense hypothesis respectively, and calculate the likelihood ratio of the prosecution hypothesis and the defense hypothesis based on the likelihood function;

[0147] The prosecution hypothesis is: the target individual and n unknown individuals are contributors to the mixed DNA sample;

[0148] The defense hypothesis is: n+1 unknown individuals are contributors to the mixed DNA sample;

[0149] S1: Obtain all possible genotype combinations at each locus;

[0150] According to the prosecution hypothesis and the defense hypothesis, the MH sequencing data of the mixed DNA sample to be tested and the typing information of the target individual, all possible genotype combinations G are listed at each locus in each hypothesis.

[0151] First, extract all unfiltered base sequences O on the specified locus l, and form a candidate allele set a = {O, Q} with an allele dropout placeholder Q. For each unknown contributor, extract two candidate alleles from set a with replacement to form its genotype. All possible genotypes of the unknown contributors are combined in the form of a Cartesian product with replacement. If there is a target individual or reference individual under this hypothesis, the genotype combination of the unknown individual that has been constructed is combined with the known genotype of the target individual or reference individual at the locus to form all possible genotype combinations G at the locus.

[0152] S2: Calculate the genotype frequency of each contributor based on the allele population frequency information;

[0153] The genotype frequency of each contributor is calculated based on the allele population frequency information and the Hardy-Weinberg equilibrium law: When the contributor is heterozygous, such as AB, its genotype frequency is 2p A p B , where p A and p B are the frequencies of alleles A and B in the population, respectively. When the contributor is homozygous, such as AA, its genotype frequency is 2p A 2 . Then the ith genotype combination G at the specified locus i The corresponding frequency P(G i |H) is the product of the genotype frequencies of all contributors at that locus.

[0154] S3: Obtain the corresponding weights based on the genotype combination and the number of MH sequence sequencing reads;

[0155] The corresponding probability or density is calculated based on all possible genotype combinations at each locus and the sequencing read information of the MH allele sequence, which is called weight.

[0156] First, the sequencing read information O of all sequences O on the specified locus l is obtained c Sum signal ratio O r .

[0157] In the specified genotype combination G i Calculate the effective number of alleles A for each sequence in set O, where the effective number of alleles of sequence o is A o :

[0158]

[0159] Where: is the number of sequences o in the sequence set O contained in the genotype of the xth contributor, M xis the mixing ratio parameter of the x-th contributor, x is the contributor number, X is the number of contributors, X=2;

[0160] According to the significant number of alleles A o Classify all sequences O, if A o If it is not 0, it is the allele sequence. n If it is 0, it is a noise sequence, and the sequence included in the genotype combination but not included in all sequence sets O is a missing allele sequence;

[0161] The weights of the allele sequence, noise sequence, and missing allele sequence are calculated separately, and then the weights are multiplied to obtain the weight corresponding to the genotype combination.

[0162] The weight of the missing allele sequence D o for:

[0163] D o =(P(Dropout)) z

[0164] Where: P (Dropout) is the allele dropout rate of 0.05, z is the number of allele dropouts in the genotype combination, the loss of one allele in a heterozygote is counted as 1 time, and the loss of a homozygote is counted as 2 times.

[0165] The weight of the noise sequence D no is the product of the Gehry density functions of each noise sequence:

[0166]

[0167] Where: k is the noise sequence number, K is the number of noise sequences, r k is the signal ratio of the kth noise sequence at the locus, r k :beta(τ1,τ2), τ1 and τ2 are noise distribution parameters, τ1 = 0.32, τ2 = 530.19, beta represents the probability density function of beta distribution;

[0168] Weight of allele sequence D a is the product of the probability density functions for each allele sequence:

[0169]

[0170] Sequencing reads y for the nth allele at a locus n Normal distribution

[0171] y n :norm(A n *α l μ,CV*A n*α l μ);

[0172] Where: α l is the specific detection performance parameter of locus l, as shown in Table 1, μ is the mean parameter of normal distribution, CV is the variability parameter of normal distribution, n is the allele sequence number, N is the number of allele sequences at the locus, and A n is the significant number of alleles of the nth allele, and norm represents the probability density function of the normal distribution.

[0173] Then at the specified locus l, the i-th genotype combination G i Weight calculation for:

[0174]

[0175] S4: Calculate the likelihood function Lik on a locus l l :

[0176]

[0177] Where: i is the genotype combination number, G i is the ith genotype combination, P(G i |H) is G i The corresponding frequency in the group, H is the prosecution hypothesis or the defense hypothesis, is the ith genotype combination G i The weight of

[0178] S5: Calculate the likelihood function Lik of all loci of the sample to be tested under the prosecution hypothesis and the defense hypothesis:

[0179]

[0180] Where l is the locus number and L is the number of all loci.

[0181] The process of calculating the likelihood ratio of the prosecution hypothesis and the defense hypothesis based on the likelihood function is as follows:

[0182] Parameters of the likelihood function under the prosecution hypothesis and the defense hypothesis are estimated respectively; the estimated parameters include: normal distribution homogeneity parameter μ, normal distribution variability parameter CV and mixing ratio parameter M of each contributor x , the number of mixing ratio parameters is the number of assumed contributors minus 1. The maximum likelihood estimation method is used to estimate the value of the parameter, and each sample is estimated three times.

[0183] According to the estimated parameter values ​​(three sets of parameter values), the size of the likelihood function is calculated under the prosecution hypothesis and the defense hypothesis respectively, and the maximum value of the likelihood function under the prosecution hypothesis and the defense hypothesis is taken respectively;

[0184] Get the likelihood ratio of the prosecution hypothesis and the defense hypothesis, the result log 10 LR = 30.4948. This result supports the hypothesis that the target individual is a contributor to the mixed sample and quantitatively reflects the strength of the evidence.

[0185] Step 5: Calculate the likelihood function values ​​of all genotype combinations at each locus, and obtain the results of the size ranking of the locus likelihood function values, as shown in Table 3. The genotype combination of the most likely contributor at each locus can be obtained as the result of mixed DNA deconvolution.

[0186] Table 3. Locus deconvolution results

[0187]

[0188] The present invention can accurately and objectively analyze the results of MH sequencing data of mixed DNA samples, reduce the uncertainty and subjective errors in the analysis process, significantly expand the range of samples that can be analyzed, and improve the accuracy of MH sequencing data analysis of mixed DNA. Since the method of the present invention does not require manual analysis, its accuracy and efficiency are significantly improved, and the efficiency of forensic analysis of MH sequencing data is improved. This improvement makes complex data processing more efficient and saves analysis time. In addition, in the analysis of mixed and complex mixed MH data, more information hidden in the sequencing data can be deeply excavated and explained.

[0189] The present invention provides a scientific method for analyzing and interpreting MH sequencing data, which not only promotes the practical application of MH genetic markers in the field of forensic medicine, but also lays a foundation for its future development.

Claims

1. A hybrid DNA probabilistic typing method for MH sequencing data, characterized in that: The following steps are involved: Step 1: construct a training set based on the MH sequencing data of a single individual's DNA, and estimate the prior parameters based on the training set; Step 2: Obtain characteristic information of MH sequencing data of the mixed DNA sample to be tested, the characteristic information includes MH typing and quantitative information of the mixed DNA, typing information of the target individual, typing information of the reference individual and related parameters; MH sequencing data of the mixed DNA sample to be tested includes: MH locus, total sequencing reads of MH locus, MH sequence base information, and sequencing reads of MH sequence; The target individual typing information includes: MH allele base information of the target individual at each MH locus; The reference individual typing information includes: MH allele base information of the reference individual at each MH locus; Related parameters include: prior parameters and number of contributors Step 3: Filter out noise sequences based on sequencing information; Step 4: Based on the prior parameters obtained in step 1 and the feature information obtained in step 2, construct likelihood functions under the prosecution hypothesis and the defense hypothesis respectively, and calculate the likelihood ratio of the prosecution hypothesis and the defense hypothesis based on the likelihood function; The prosecution hypothesis is: the target individual and n unknown individuals are contributors to the mixed DNA sample, or the target individual, the reference individual, and n-1 unknown individuals are contributors to the mixed DNA sample; The defense hypothesis is: n+1 unknown individuals are contributors to the mixed DNA sample, or the reference individual and n unknown individuals are contributors to the mixed sample; The target individual is the individual suspected to be one of the contributors to the mixed DNA sample; A reference individual is an individual that has been identified as one of the contributors to a mixed DNA sample; The likelihood function is constructed as follows: S1: Obtain all possible genotype combinations at each locus; S2: Calculate the genotype frequency of each contributor based on the allele population frequency information; S3: Obtain the corresponding weights based on the genotype combination and the number of MH sequence sequencing reads; The process of obtaining the corresponding weights according to the genotype combination and the MH sequence sequencing reads is as follows: Get the sequencing read information O of the set O of all sequences at the specified locus l c Sum signal ratio O r ; In the specified genotype combination G i Calculate the effective number of alleles A for each sequence in set O, where the effective number of alleles of sequence o is A o Where: B ox is the number of sequences o in the sequence set O contained in the genotype of the xth contributor, M x is the mixing ratio parameter of the x-th contributor, x is the contributor number, and X is the number of contributors; According to the significant number of alleles A o Classify all sequences. If A o If it is not 0, it is the allele sequence. o If it is 0, it is a noise sequence, and the sequence included in the genotype combination but not included in all sequence sets O is a missing allele sequence; Calculate the weights of the allele sequence, noise sequence, and missing allele sequence respectively, and then multiply the weights to get the weight corresponding to the genotype combination; S4: Calculate the likelihood function Lik on a locus l l : Where: i is the genotype combination number, G i is the ith genotype combination, P(G i |H) is G i The corresponding frequency in the group, H is the prosecution hypothesis or the defense hypothesis, is the ith genotype combination G i The weight of S5: Calculate the likelihood function Lik of all loci of the sample to be tested: In the formula, l is the locus number, and L is the number of all loci; Step 5: Calculate the likelihood function values ​​of all possible genotype combinations at each locus and obtain the result of ranking the likelihood function values ​​of the loci.

2. A hybrid DNA probabilistic typing method for MH sequencing data according to claim 1, characterized in that: The priori parameters in step 1 include: locus-specific detection performance parameters of MH genetic markers, analysis threshold, allele dropout rate, allele population frequency, and noise distribution parameters.

3. A hybrid DNA probabilistic typing method for MH sequencing data according to claim 1, characterized in that: The process of filtering out the noise sequence in step 3 is as follows: According to the MH sequencing data of the mixed DNA sample to be tested, the signal ratio r of each base sequence is calculated; The MH sequencing data of the mixed DNA sample to be tested is filtered according to the analysis threshold, and those that are not within the threshold range are filtered out.

4. A hybrid DNA probabilistic typing method for MH sequencing data according to claim 1, characterized in that: The process of calculating the genotype frequency of each contributor in S2 is as follows: The genotype frequency of each contributor was calculated based on the allele population frequency information and the Hardy-Weinberg equilibrium law.

5. A hybrid DNA probabilistic typing method for MH sequencing data according to claim 1, characterized in that: The process of calculating the likelihood ratio of the prosecution hypothesis and the defense hypothesis according to the likelihood function in step 4 is as follows: Parameters of the likelihood function under the prosecution hypothesis and the defense hypothesis are estimated respectively; According to the estimated parameter values, the size of the likelihood function is calculated under the prosecution hypothesis and the defense hypothesis respectively, and the maximum value of the likelihood function under the prosecution hypothesis and the defense hypothesis is taken respectively; Get the likelihood ratio between the prosecution hypothesis and the defense hypothesis.

6. A hybrid DNA probabilistic typing method for MH sequencing data according to claim 5, characterized in that: The estimated parameters include: normal distribution mean parameter, normal distribution variability parameter, and mixing ratio parameter for each contributor.

7. A hybrid DNA probabilistic typing method for MH sequencing data according to claim 1, characterized in that: The weight D of the missing allele sequence o for: D o =(P(Dropout)) z Where: P (Dropout) is the allele dropout rate, z is the number of allele dropouts in the genotype combination; The weight of the noise sequence D no for: Where: k is the noise sequence number, K is the number of noise sequences, r k is the signal ratio of the kth noise sequence at the locus, and are noise distribution parameters, beta represents the probability density function of beta distribution; Weight of allele sequence D a for: Sequencing reads y for the nth allele at a locus n Normal distribution y n ~norm(A n *a l m, CV*A n *a l m) Where: α l is the specific detection efficiency parameter of locus l, μ is the mean parameter of normal distribution, CV is the variability parameter of normal distribution, n is the allele sequence number, N is the number of allele sequences at the locus, and A n is the significant number of alleles of the nth allele, and norm represents the probability density function of the normal distribution.

Images