Evaluation of microhaplotype panels for complex kinship analysis using massively parallel sequencing
Jiaming Xue,Mengyu Tan,Ranran Zhang,Dezhi Chen,Guihong Liu,Yazi Zheng,Qiushuo Wu,Yuanyuan Xiao,Miao Liao,Shengqiu Qu,Weibo Liang
DOI: https://doi.org/10.1016/j.fsigen.2023.102887
2023-05-04
Abstract:In recent years, microhaplotypes (MHs) have become a research hotspot within the field of forensic genetics. Traditional MHs contain only SNPs that are closely linked within short fragments. Herein, we broaden the concept of general MHs to include short InDels. Complex kinship identification plays an important role in disaster victim identification and criminal investigations. For distant relatives (e.g., 3 rd -degree), many genetic markers are required to enhance power of kinship testing. We performed genome-wide screening for new MH markers composed of two or more variants (InDel or SNP) within 220 bp based on the Chinese Southern Han from the 1000 Genomes Project. An NGS-based 67plex MH panel (Panel B) was successfully developed, and 124 unrelated individual samples were sequenced to obtain population genetic data, including alleles and allele frequencies. Of the 67 genetic markers, 65 MHs were, as far as we know, newly discovered, and 32 MHs had effective number of allele (A e ) values greater than 5.0. The average A e and heterozygosity of the panel were 5.34 and 0.7352, respectively. Next, 53 MHs from a previous study were collected as Panel A (average A e of 7.43), and Panel C with 87 MHs (average A e of 7.02) was formed by combining Panels A and B. We investigated the utility of these three panels in kinship analysis (parent-child, full siblings, 2 nd -degree, 3 rd -degree, 4 th -degree, and 5 th -degree relatives), with Panel C exhibiting better performance than the two other panels. Panel C was able to separate parent-child, full-sibling, and 2 nd -degree relative duos from unrelated controls in real pedigree data, with a small false testing level (FTL) of 0.11% in simulated 2 nd -degree duos. For more distant relationships, the FTL was much higher: 8.99% for 3 rd -degree, 35.46% for 4 th -degree, and 61.55% for 5 th -degree. When a carefully chosen extra relative was known, this may enhance the testing power for distant kinship analysis. Two twins from the Q family (2-5 and 2-7) and W family (3-18 and 3-19) shared the same genotypes in all tested MHs, which led to the incorrect conclusion that an uncle-nephew duo was classified as a parent-child duo. In addition, Panel C showed great capacity for excluding close relatives (2 nd -degree and 3 rd -degree relatives) during paternity tests. Among 18,246 real and 10,000 simulated unrelated pairs, none were misinterpreted as a relative within 2 nd -degree at a log 10 (LR) cutoff of 4. The panels presented herein could provide supplementary power for the analysis of complex kinship.
genetics & heredity,medicine, legal