The Effects of Heat and Explosions on Forensic DNA Analyses

This project explores the effects of high temperatures and explosions on DNA samples of forensic value. It aims to determine the conditions under which biomaterials may degrade in accidental disasters, crime involving fire, and bombings, and how degradation affects interpretation. The experiments were designed to reflect situations requiring victim and suspect identification, and so involved testing various human samples including blood and saliva stains. Some samples were heated under laboratory conditions, while others were attached to pipe bombs and detonated outdoors with the assistance of US Police bomb enforcement officers. A sensitive mitochondrial DNA multiplex system was devised, and successfully used to detect DNA degradation prior to more costly analyses. Capillary electrophoresis-based STR typing and Massively Parallel Sequencing were compared in terms of their performance on degraded DNA. Treatment at 180 °C for 30 minutes was required to induce the first signs of DNA degradation in dried blood and saliva stains, reflected by reduced post-PCR DNA detection or drop-out of longer amplicons. There were no interpretable DNA products when heat treatment increased to 200 °C. Similar degraded DNA effects were observed in 27% of stains placed on and within smokelesspowder-charged pipe bombs, but no sign of degradation was observed with the more energetic C4 explosive, probably because of the shorter duration of heat exposure. DNA degradation poses challenges to the interpretation of retrieved genetic data. These challenges were investigated both in real profiles from heat-treated samples, and in simulated data. In addition, a real case of an unidentified male victim (the Blazing Car Murder 1931) was analysed. The combination of real and simulated data provided realistic scenarios, but also allowed control of parameters which affect evidential strength such as population size and diversity, through lowering of the likelihood ratio and increasing the number of random matches in a database.