A likelihood ratio framework for inferring close kinship from dynamically selected SNPs.

Forensic genetic genealogy (FGG) is a force-multiplier for human identification, leveraging dense single nucleotide polymorphism (SNP) data to infer relationships through identity by descent (IBD) segment analysis. Although powerful for investigative lead generation, broad adoption of SNP-based identification methods by the forensic community, especially medical examiners and crime laboratories, necessitates likelihood ratio (LR)-based relationship testing, to align with traditional kinship testing standards. To address this gap, a novel method was developed that incorporates LR calculations into FGG and SNP testing workflows. This approach is unique in that it dynamically selects unlinked, highly informative SNPs based on configurable thresholds for minor allele frequency (MAF) and minimum genetic distance for a robust and reliable analysis. Employing a curated panel of 222,366 SNPs from gnomAD v4 and data from the 1,000 genomes project, high accuracy in resolving relationships up to second-degree relatives can be achieved. For example, a subset of 126 SNPs (MAF > 0.4, minimum genetic distance of 30 cM) yielded 96.8% accuracy and a weighted F1 score of 0.975 across 2,244 tested pairs. This LR-based methodology enables forensic laboratories to select informative SNPs and integrate modern genomic data with existing accredited relationship testing frameworks, providing critical statistical support for close-relationship comparisons and enhances the rigor of FGG- and SNP-based human identification applications.