Likelihood evaluation using 15 common short tandem repeat loci: a practical and simulated approach to establishing...

In individual identification, often the only way to circumvent the lack of personal reference materials is to resort to a kinship analysis between the unidentified party (UP) and putative relatives.

Genotypes of 15 short tandem repeat loci in 20,000 families were simulated and the likelihood ratios (LRs) were calculated based on the frequencies of alleles in the Japanese population. These LR values illustrate the likelihood of whether an alleged kin is related to a UP. First, three sets of kinship were assessed: UP and one parent (or child); UP and two children; and UP and both parents. Next, an UP was compared with one to three alleged siblings (ASs). In the end, these calculations were applied to 21 actual cases where the identity of an individual lay in dispute.

When comparisons were made between a UP and both parents, or two children, the LR median dramatically increased (6.68 x 10(11) and 1.10 x 10(8), respectively) in comparison to assessments involving an UP and one parent or child (48,900). Similarly, when two or three true siblings were made available for comparison, the LR was far greater (7.32 x 10(6) and 4.50 x 10(8), respectively) than in cases where only one sibling was available (9,250). Of the 21 actual cases these calculations were applied to, 1 case ended in exclusion with absolute certainty and the remaining 20 cases obtained high LRs.

As demonstrated, this system is extremely effective in cases of personal identification where both parents, two or more children, or two or more ASs are available for comparison.