Estimating Number of Contributors in Massively Parallel Sequencing Data of STR loci

Publication date: Available online 26 September 2018Source: Forensic Science International: GeneticsAuthor(s): Brian A Young, Katherine Butler Gettings, Bruce McCord, Peter M. ValloneAbstractIn recent years a number of computer-based algorithms have been developed for the deconvolution of complex DNA mixtures in forensic science. These procedures utilize likelihood ratios that quantify the evidence for a hypothesis for the presence of a suspect in a DNA profile compared to an alternative hypothesis. Proper operation of these software systems requires an assumption regarding the total number of contributors present in the mixture. Unfortunately, estimates based on counting the number of alleles at a locus can be inaccurate due to the sharing and masking of alleles at individual loci. The effects of allele masking become increasingly severe as the number of contributors increases, rendering estimates about high-order mixtures uncertain. The accuracy of these estimates can be improved by increasing the number of STR markers in panels, and by using highly polymorphic markers. Increasing the number of STR markers from 13 to 20 (expanded CODIS panel) improves the accuracy of allele count-based estimation methods for low-order mixtures, but accuracy for high-order mixtures (> 3 contributors) remains poor due to allele masking. An alternative technique, massively parallel sequencing, holds great potential to improve the accuracy of the estimate of number of contributors due to its ab...
Source: Forensic Science International: Genetics - Category: Forensic Medicine Source Type: research