Likelihood-Ratio

🧬 Introduction In humanitarian and forensic genetics, one of the most computationally tedious problems is also one of the most consequential: systematically comparing a set of reference family pedigrees against a set of candidate biological relatives to find which pairs are actually related to the same missing person. This is the core challenge of Missing Person Identification (MPI) and Disaster Victim Identification (DVI) workflows. A defining constraint in all these scenarios is that the person of interest is not always genotyped. They may be deceased and unidentified, disappeared, suspected of theft and identity swap, or simply unwilling to provide a biological sample — or their remains may have been cremated or commingled in an ossuary, making individual sampling no longer possible. The analysis must therefore rely exclusively on the genotypes of their relatives — making indirect kinship inference the only available path to identification. ...

🧬 Introduction In forensic genetics, Short Tandem Repeats (STRs) have been the key of human identification for over two decades. Current forensic systems typically use 24 autosomal STR markers, forming the global standard for paternity testing, missing person identification (MPI) and disaster victim identification (DVI). However, the field is now gradually migrating toward high-density Single Nucleotide Polymorphism (SNP) technologies. This transition is motivated by the need for greater statistical resolution, improved ability to detect distant kinship, and compatibility with degraded DNA samples often recovered in humanitarian and forensic contexts. ...