Study On Forensic SNP Kinship Inference Based On High-throughput Sequencing

Currently,the conventional genetic marker short tandem repeat(STR)in forensic DNA field can be used for kinship analysis.Autosomal STR can be used for kinship analysis within two degrees of kinship such as parent-child and full siblings.Y-STR can be used for paternal kinship analysis but cannot determine the degree of kinship.To improve the ability of longdistance kinship analysis,this study established a forensic SNP pedigree inference technology based on high-throughput sequencing technology.First,a pedigree inference technology based on whole-genome sequencing was established,which can analyze 1-7 degrees of kinship.Then,a capture sequencing system of 9kSNP was constructed(hereinafter referred to as 9kSNP),which can analyze 1-5 degrees of kinship.The capture success rate,probe uniformity,typing accuracy and pedigree inference accuracy of the locus set were systematically studied.This study provides detection technology and reagent support for the forensic SNP pedigree inference technology in forensic medicine application,promoting the more effective service of forensic DNA technology in case investigation.Firstly,this study explored the feasibility of whole-genome sequencing technology for forensic pedigree inference and established a forensic SNP pedigree inference whole-genome sequencing system based on domestic sequencing platforms.Secondly,9kSNP was screened from the high-density SNP locus set.In addition,9858 locus sets were selected from the mature systems of 120AI-SNP,145Y-SNP,57II-SNP,43PI-SNP,4ABO-SNP,2Amel,82 mt SNP and426 r CRS in the laboratory previous research.Corresponding ss DNA liquid hybridization capture probes were designed and synthesized.A sequencing reagent system was constructed through the Huada MGISEQ-200 RS sequencing platform.XGBoost algorithm was used to predict kinship relationships.A series of evaluations were conducted on typing accuracy and pedigree inference accuracy to establish a forensic SNP capture sequencing and pedigree inference whole process based on domestic sequencing platforms.The results of this study showed that the consistency rate between whole genome sequencing SNP typing and Wegene GSA chip typing was above 99.62%.The IBS algorithm can be used to predict kinship relationships from level 1 to level 4,and the IBD algorithm can be used to predict kinship relationships from level 1 to level 7.The accuracy of predicting kinship relationships at level 7 is 100%.The pedigree inference ability of SNPs obtained through high-depth whole-genome sequencing data is not significantly different from that of chip prediction results.At the same time,the results of population inference and investigation using whole genome sequencing data are consistent.The 9kSNP is evenly distributed on human autosomes,and the confidence interval accuracy for four-level kinship prediction using XGBoost algorithm is 98.5%,with no false negative prediction relationship.The capture sequencing system based on domestic MGISEQ-200 RS platform can capture target fragments with all locus probes.At most,only two locus probes in standard samples,venous blood and saliva DNA failed to capture.The proportion of loci with depth greater than one times the average depth in each sample is about 50%,and the uniformity of probes is excellent.The consistency rate between 9kSNP sequencing results and Wegene CGA chip typing is above99.84%,and the consistency rate between standard sample capture sequencing results and1000 G database typing is above 99.76%.Even when the starting DNA is 5.04 ng,a consistency rate of 99.85% can still be obtained,indicating that the system has low requirements for sample DNA quantity.The capture sequencing system can infer kinship relationships up to level 5,with a false negative rate of 0 for kinship relationships up to level 5,which will not cause loss of relationship pairs.False negatives appear at level 6,but they can still provide some reference clues for case solving.The functional loci in the integrated detection system can accurately predict ancestors,pigment phenotypes and haplogroups,and the mitochondrial maternal investigation results can complement the kinship inference results of 9kSNP,which is more helpful for practical application of forensic pedigree inference.