Kinship analysis: assessment of related vs unrelated based on defined pedigrees.

Abstract

The study aimed at evaluating whether the adoption of enlarged batteries of STR markers in kinship analysis may provide LR values suitable for discrimination of relatives from non-relatives, in comparison to conventionally used STR panels. The presence of LD among some loci and its effects on LR values were also assessed. Three hundred pairs of related and unrelated individuals, each separated from 1-3 generations and residing in North Italy were genotyped with the Investigator HDplex STR kit (Qiagen), AmpFlSTR Identifiler (Applied Biosystems), and PowerPlex Fusion System (Promega). Loci and alleles shared between each pair and within groups of relatives were compared. Also, combined LR values with and without loci in LD, sensitivity and specificity were calculated for each commercial kit and their combinations. Full siblings displayed the largest number of shared loci and alleles, with a proportion of LR ≥ 10 results significantly higher than other degrees of relatedness and, consequently, with the lowest percentage of inconclusive and false negative results. Only minor differences were detected in the combined LR distributions, after including or omitting loci in LD. However, these became only appreciable when analyzing more distant relative pairs.The implementation of additional STRs into the LR calculation allowed a complete and robust discrimination between relatives and non-relatives only for full siblings, by removing the typical uncertainty of the "grey zone", while this was not achieved among other degrees of relatedness. Furthermore, the presence of loci in LD seems to not significantly affect LR distributions within each generation.