Forensically Relevant Applications of Genome-Wide Association Studies

Abstract

The past century has seen great improvements in the science underpinning forensic investigations. The field has advanced from Lombardo's classification of the "criminal" physique, to the automated fingerprint matching systems. At the biological level, the field has moved from identification based on blood groups to automated genotyping, which is used to compare DNA extracted from saliva or blood left at a crime scene to reference DNA collected from individuals.In a legal context, however, such improvements are of limited use if relevant parties cannot fully and accurately understand the evidence. This includes both sides of legal representation and, where a jury is involved, eligible jury members. Previous studies have examined the understanding of statistical evidence amongst jury-eligible individuals, finding that they commonly result in logical fallacies. These types of studies have typically focussed on the understanding of DNA matching methods. However, genomic analyses currently focus on the use of genome-wide association analyses (GWAS) which aim to identify single-base variants associated with the trait being studied. Many forensically relevant traits have been studied using GWAS approaches and there is substantial interest in the use of these findings to predict phenotypes of interest, and the implications of using these findings as supporting evidence during trials.While there have been a number of GWAS studies of eye and hair colour and hair morphology, there are some notable gaps in current literature on genetic variants influencing forensically relevant complex traits. Although fingerprint patterns are commonly used for identification, there have been no studies seeking to identify genetic variants influencing these patterns. Similarly, although there are a number of papers focussing on gross hair morphology, there have been no studies on the micro-level morphology. In order to address these limitations, the current thesis aims to identify genetic variants that contribute to differences in fingerprint patterns, quantify the genetic contribution to and identify genetic variants influencing micro-level hair diameter and curvature.In addition to the identification of genetic variants influencing forensically relevant traits, human factors are also of interest. In order to introduce GWAS evidence in a comprehensible manner to a legal and general audience, it is apparent that clear communication of scientific ii and statistical evidence is essential. By studying attitudes and perception of GWAS evidence by a jury-eligible population, this thesis provides the first step towards an evidence-based approach to transmit knowledge reliably in a forensic context.