Detection of minor DNA components in biological mixtures has increased as molecular techniques have become more sensitive, and thus, is an ongoing topic of concern in the forensic DNA community. A microhaplotype-based sequencing assay and analysis methodology offers multiple advantages for complex and imbalanced mixtures that address inherent issues of current short tandem repeat (STR) methods. For example, the increased power of discrimination, elimination of stutter artifacts, and reduction of allelic or size-specific amplification bias exhibited by microhaplotypes will enhance mixture deconvolution by improving detection of minor contributors and improving the certainty of genotype deconvolution. Forty-three microhaplotype loci were identified based on high effective number of Alleles (Ae) values as a predictor for their efficacy in mixture deconvolution and a target amplification custom assay was constructed using AmpliSeq for Illumina chemistry with MiSeq FGx sequencing. The assay was evaluated for sensitivity, inhibition effects, and mixture analysis using in vitro constructed mixtures and mock transfer mixtures collected from gun crime-relevant substrates. A total of 189 mixtures were constructed consisting of 2 - 5 contributors with minor contributor proportions ≥1%. Additionally, related contributors were included in the mixtures and the POI comparison database to assess deconvolution in the presence of familial allele sharing. Then, mixture deconvolution was performed with two probabilistic genotyping methods: NexGenID, a novel software platform optimized for mixture deconvolution and probabilistic genotyping of microhaplotype sequence data, and EuroForMix, a widely used open-source probabilistic genotyping software modifiable for use with microhaplotype sequence data.
The constructed microhaplotype panel demonstrated high discriminatory power with combined match probabilities ranging from 9.53E-52 to 4.79E-63, the ability to infer biogeographical ancestry, sensitivity down to 50 pg inputs, and applicability to inhibited or degraded trace samples. Application to complex DNA mixture samples demonstrated the assay’s potential to exceed minor-contributor detection when compared to STR deconvolution with STRmix. True contributors generated Log(LR)s >0 for all mixtures deconvolved using NexGenID and EuroForMix, and no non-contributors generated Log(LR)s >0. Additionally, instances were observed of minor contributors to 3-person, 4-person, and 5-person mixtures where the STR Log(LR) gave weak support or even no support (Log(LR)<0) for inclusion of the profile to the mixture, but the ability to capture more alleles with greater allelic diversity using microhaplotypes produced Log(LR)s >10. These findings underscore that as massively parallel sequencing is adopted in forensic DNA laboratories, microhaplotypes are well positioned to integrate seamlessly alongside STRs rather than requiring an entirely new interpretive ecosystem.