Practical Implementation of ForenSeq™ DNA Signature Sequence-Based Mixture Interpretation, Sample Comparison and Populations Statistics Methods for Criminal Casework

Summary

Background: Traditional forensic DNA analysis for short tandem repeats (STRs) using capillary electrophoresis (CE) has been the gold standard for many years in the community. Newer technologies such as massive parallel sequencing (MPS) offer several advantages over the traditional STR workflow including the ability to combine multiple marker types like STRs and single nucleotide polymorphisms (SNPs) into a single assay to generate more data from less sample. Additionally, MPS offers the ability to generate sequences of STRs to increase discriminatory power for identity and lineage analysis as well as contribute to mixture detection and interpretation. As a significant proportion of forensic DNA samples are mixtures, the ability to reliably interpret mixtures is a substantial challenge for forensic laboratories. Mixture validation studies are increasingly important for the implementation of sequencing in a forensic setting. We present a comprehensive mixture validation study for the ForenSeq™ DNA Signature Prep (Sig Prep) using DNA Primer Mix B (DPMB) kit sequenced on the MiSeq FGx® Sequencing System and Universal Analysis Software v1.3 (UAS) with the goal of validating MPS autosomal STR (auSTR) mixture interpretation for two to four contributor mixtures as well as the reliable deduction of major contributors.

Methods: The Sig Prep kit includes 231 marker types across auSTRs, Y-STRs, X-STRs, identity SNPs, ancestry SNPs, phenotype SNPs, and Amelogenin for robust characterization of samples with 1 ng of DNA input. This validation study focused on the 26 auSTR markers (locus D22S1045 was excluded). The additional markers will be included in subsequent studies. A total of 44 mixtures spanning a range of mixture ratios, DNA inputs, and contributors were created with total DNA inputs ranging from 50.5 ng to 0.025 ng. Mixtures were processed in duplicate along with two replicates of the NIST Standard Reference Material 2391d Component D across 3 sequencing runs for a total of 90 mixtures sequenced. The mixtures were also processed with the GlobalFiler™ PCR Amplification Kit (GF) and typed on a 3500xL Genetic Analyzer. The mixtures were analyzed in the UAS and custom Excel workbooks. For each mixture, loci with complete donor resolution were evaluated to determine the number of contributors, observed mixture ratios, read counts for the major and minor contributor, and intralocus balance of the major and minor contributors across duplicate amplifications and mixture series. The ability to deduce a major profile was assessed and proposed rules for manual mixture interpretation were compiled to include loci without complete separation of alleles. Concordance between CE and MPS mixtures were evaluated.

Results: The inclusion of additional loci in the Sig Prep kit as well as the ability to differentiate shared length alleles by sequence aided in the ability to recognize mixtures and separate contributors with up to four contributors. Mixture ratios were relatively consistent within the mixtures across all loci, DNA inputs, and technologies. Reportable major profiles (minimum of 8 complete loci) were able to be deduced from two- to four-person mixtures with calculated mixture ratios of 1:3 or higher. Replicate amplifications were beneficial for reporting concordant major profiles due to Sig Prep’s sensitivity to stochastic effects like elevated stutter and imbalance, particularly at total DNA inputs of less than 0.25 ng.

Conclusions: This mixture validation study supports the use of the Sig Prep DPMB for forensic mixture analysis. Deduction of major contributors from mixtures of up to four contributors were consistent between Sig Prep and GF mixtures with Sig Prep yielding more loci per major due to additional kit loci and sequence-based alleles. Replicate amplifications of mixtures is recommended to confirm amplification variability and reliable reporting of contributor profiles. Additional analysis to be performed includes the remaining X-STR and Y-STR loci as well as all SNP markers. The basis of this study will also be used to evaluate the use of STRmix NGS R&V within the laboratory’s Sig Prep workflow.