Assessing the FBI's Native American STR database for random match probability calculations

In forensic statistics, the random match probability (RMP) is the probability that a “match” would occur by coincidence while the likelihood ratio (LR) describes the strength of DNA evidence. Using these statistics to assess the weight of DNA evidence requires an appropriate and well-characterized population-specific short tandem repeat (STR) database to reliably estimate allele frequencies. This study compared several Native American-specific STR datasets, including those not represented in the CODIS Native American database, and revealed that increasing the number of STR markers resulted in lower RMP values while a θ adjustment from 0.03 to 0.04 generated increases in RMP. To prevent biases that may arise from the underrepresentation of tribes in the current CODIS Native American database, data derived from tribes in different geographic regions and language families are necessary to ensure inclusive representation of the Native American population and generate more reliable statistical results.