Effect of decrease in the number of loci on full sibling analysis

DNA degradation or low DNA quantity (e.g., in the case of DNA extracted from bone samples) is likely to decrease the number of loci detected in short tandem repeat (STR) typing. In this study, we evaluated the effect of the decreased number of loci on full sibling analysis through the likelihood ratio (LR) calculation. Simulations were performed by decreasing the number of loci using the previously studied data for full siblings (10,000 full sibling pairs and 10,000 unrelated pairs) that had been generated for the 21 autosomal STR loci in the GlobalFiler kit using the allele frequencies of the Japanese population. For each pair, we excluded the loci one by one in order of longer to shorter fragment size (fragment size model, assuming DNA degradation) or randomly (random model, assuming low-template DNA) from the LR calculation. As a result, the effect of decreased locus number in the fragment size model was slightly larger than that in the random model. Although decrease in the number of loci used in the LR calculation reduced the distance between the LR distribution in full sibling pairs and that in unrelated pairs, it was shown that even if the number of loci decreased, this does not immediately eliminate the usefulness of the LR for full sibling analysis.