Pairwise kinship testing with a combination of STR and SNP loci

Abstract

Individual commercially available kits exhibit limited discrimination power in full-sibling and second-degree kinship analysis, and therefore they are commonly combined with other kits to obtain more loci and a higher efficacy. However, few studies have systematically evaluated the discrimination power of combined loci. In this study, we combined the ForenSeq™ DNA Signature kit (containing 27 short tandem repeats [STRs] + 91 single nucleotide polymorphisms [SNPs]) with the AGCU NC 21 + 1 PCR amplification kit (containing 21 STRs) to obtain a non-overlapping set of 40 STR and 91 SNP markers. The discrimination power was evaluated for 74 full-sibling pairs, 114 uncle/aunt-nephew/niece pairs and 93 grandparent-grandson/granddaughter pairs. The results show that the efficacy of the 40 STR + 91 SNP combination is higher than the efficacy of either 27 STRs + 91 SNPs or 40 STRs alone. Both the sensitivity and specificity of the 40 STR + 91 SNP marker set achieved 100 % in full-sibling testing, with strong power to distinguish second-degree relatives from unrelated pairs. The 40 STR + 91 SNP set could also distinguish most full-sibling relatives from second-degree relatives but was insufficient to distinguish relatives who belong to the same autosomal kinship class. Our results suggest that ignoring linkage can lead to incorrect likelihood ratios for both related and unrelated pairs, while mutation had a relatively lower effect on the likelihood ratios. Moreover, linkage and mutation had a higher impact on full-sibling testing than on second-degree kinship testing. The discrimination power of the 40 STR and 91 SNP marker set could be strengthened by adding an additional relative.