"How convincing is a matching Y-chromosome profile?" - new paper co-authored by David Balding published in PLoS Genetics
6 November 2017
Y-chromosome DNA profiles are important in forensic science, particularly when a male has been accused of assaulting a female.
However, unlike for autosomal profiles, the problem of evaluating weight-of-evidence for Y profiles has not been satisfactorily resolved despite many attempts. The key idea missing from current approaches is that Y-profile matches are due to patrilineal relatedness that is typically too remote to be recognized, but close compared with the relatedness of random pairs from the population. We focus on approximating the number of males with matching Y profiles, rather than a population fraction or match probability. We describe a simulation model of Y profile evolution, implemented in open-source software, for approximating the number of males sharing a Y profile. We extend our simulation method to also model database selection. Even under the optimistic assumption that the database has been sampled randomly in the relevant population, we show that database counts don’t help much. The reason is that modern profiling kits with high profile mutation rates imply that almost all profiles are rare relative to typical database sizes. We discuss how to use our approach to present evidence in court in a way that is both fair and easy to understand.
Read full paper: How convincing is a matching Y-chromosome profile?
Authors: Mikkel M. Andersen, David J. Balding