What started as a PhD rotation turned into a wonderful collaboration with Noah Rosenberg ✨. (4/4)

We use phased SNP-STR reference data from 2504 individuals in the 1000 Genomes Project, with SNPs surrounding the CODIS STRs. When selecting SNPs based on characteristics such as MAF and distance to the STR, we find that ~1000 SNPs suffice for record-matching, a 10-fold reduction. (3/4)

Prior work has used around 10,000 SNPs for record-matching. In this project, we asked whether we can further reduce the size of SNP panels required for record-matching with STR profiles? Reducing the number of SNPs has important implications for privacy and cost-effectiveness. (2/4)

Existing systems for forensic identification rely on STRs, but future systems may eventually rely on SNP panels. Genetic record-matching could help this transition by enabling the matching of STR and SNP profiles using linkage disequilibrium between SNPs and STRs. (1/4)

Huge congratulations!! So so excited for your lab!