Single step strategy for estimating unknown STR allele frequencies in a population

Abstract

Constant growth of forensic DNA databases requires new STR markers to be developed, tested and validated. Aiming to simplify the process of obtaining allele frequency distributions of new STR loci included in new megaplexes, a single step protocol was developed. This goal was achieved by pooling samples of genetically unrelated donors, previously genotyped (n = 99), containing the same amount of DNA (total DNA = 4.5 ng/ul) as measured by fluorometry. Suitability of the method was tested by amplifying commercial kits: GlobalFiler (GF), Verifiler Express (VFE) and Y-filer plus (YFP). Allele peak heights was used to estimate allele frequency and compared with the actual frequency calculated by counting method of the individual pooled samples. The multinomial goodness of fit test was performed for testing statistical differences.In case of VFE, 4.5 ng was the DNA amount that could best estimate the allele frequency distribution; 1 ng for GF and YFP. For simple, complex and compound makers, the highest allele frequencies were correctly estimated (p > 0.05) but in some cases, it was not able to detect alleles in very low frequencies, 2/198 or below, either with VFE or GF.In case of D12S391, either for VFE and GF, some microvariants with frequency of 2, 3 or 5/198 were not assigned although they were present in the profiles, probably due to POP7 resolution. In case of YFP, some alleles were not detected at DYS385.Nevertheless, with some limitations, the proposed strategy proved to be fast and efficient for major allele frequency estimation. This simple approach generates an empirical distribution that allows for the estimation of population allele frequencies in a single step assay.