| Abstract: | Almost half of all US households own a dog (Canis familiaris). Though these household pets can attack humans and other animals, they are also frequently victims of cruelty, neglect and theft. In human‐oriented investigations, the tendency of domestic dogs to leave behind physical traces (such as hair) can serve as valuable links between crime scenes and suspects/victims. This demonstrated utility of canine biological evidence has created demand for genotyping marker sets for canine forensic genetic testing. Through research and casework, short tandem repeat (STR) panels have been shown to be very efficient for identity and parentage testing in dogs. However, there is an absence of comparative studies between different canine forensic identification kits. The Thermo Fisher Scientific Canine Genotypes ? Panel 1.1 and 2.1 Kits were originally designed and developed for routine and forensic use respectively, although both kits can be used for either purpose. In this study, we evaluated both STR panels to determine how critical forensic genetic metrics are affected by panel‐to‐panel variation in marker composition and design. Our results show that although STR panel composition can influence estimates such as inbreeding, combined power of discrimination and combined probability of exclusion, greater average allele number values exhibited across all markers in Panel 2.1 facilitated significantly more precise estimates of random match probability (RMP) and combined probability of identity. Furthermore, we demonstrate that a theta (θ) correction of 0.09 can be used to conservatively adjust RMPs generated from a small reference database of fewer than 50 samples, confirming that Panel 2.1 is a more robust forensic genotyping system than is Panel 1.1. for domestic dogs. We also demonstrate that opportunistic local sampling of fewer than 50 mixed‐breed dogs can produce sufficient discriminatory and exclusionary power with either genotyping kit. |
