SUMMARIZING DNA EVIDENCE WHEN RELATIVES ARE POSSIBLE SUSPECTS

Summaries of deoxyribonucleic acid (DNA) evidence in criminal proceedi ngs have been controversial in part because calculations of the probab ility of a ''match'' between crime and suspect samples assuming that t he pairing of these samples occurred by chance typically have depended on independence or conditional independence assumptions that are subj ect to dispute. This article describes a new methodology that summariz es DNA evidence by addressing the possibility that a relative of tile accused individual is the source of a crime sample, which can be viewe d as complementary to assessing the probability of a ''match'' assumin g that individuals in a broader population are paired by chance. The n ew technique uses a statistical criterion based on distances between b ands from autoradiograph images to distinguish individuals without rel ying on any conditional independence assumptions. We apply the method to data from a paternity testing laboratory to illustrate its ability to distinguish between the DNA profiles of related individuals. In add ition, a simulation study suggests that to distinguish reliably betwee n family members, more genetic loci should be probed than the three to six loci typically investigated in current practice.