Combining evidence across multiple mixed DNA profiles for improved resolution of a donor when a common contributor can be assumed

Abstract

In casework there are often multiple mixed DNA profiles with a single unknown offender who is believed to be a contributor. Sometimes none of these profiles are individually informative enough to evaluate whether a person of interest (PoI) is a contributor. We propose a method that combines evidence across multiple mixtures to better resolve the genotype of a queried common donor. This method can also resolve genotypes of a queried non-common donor (i.e. unique to one profile), when other donors of that mixture have contributed to multiple samples. The approach has similarities with the analysis of replicate PCRs, with the difference being that we do not necessarily assume that all the contributors are the same across the evidential profiles, nor do we assume that parameters such as the mixture ratio are constant. Our method can be used to compute an LR for a PoI being the common contributor to multiple stains. It is also possible to interrogate a database of reference profiles to search for the queried donor (common or non-common). We show that the method can identify a queried contributor in cases where individual comparisons have limited capacity to discriminate between donors and non-donors, when some assumption(s) can be made about multiple contributors being from the same sources (queried or not).

Keywords: DNA intelligence; common contributor; deconvolution; forensic DNA; genotype resolution.