UCL Genetics Institute, Research Department of Genetics, Evolution and Environment, London, & Cellmark Forensic Services, Abingdon.
A PhD studentship in Statistical Genetics, starting October 2012, is available in the University College London (UCL) Genetics Institute (UGI, http://www.ucl.ac.uk/ugi/).
Forensic DNA evidence has had a huge impact on crime detection in the 20 years since it was first introduced, but it has also been a topic of considerable controversy surrounding for example population genetics issues related to match probabilities across ethnic groups. In recent years enhanced techniques have been introduced to recover and profile DNA samples, down to just a few tens of picograms, or with substantial levels of degradation and contamination. The evaluation of such 'noisy' profiles has also been controversial, for example in the case R v Hoey in the UK or that of Knox and Sollecito in Italy. The student will help develop, assess and apply computational techniques for robust interpretation of the "noisy" results of such profiling, which include stochastic effects such as allele dropout and dropin, exaggerated stutter and other artefacts. Mixtures of DNA from multiple sources, some of them at very low level, frequently occur and these stochastic effects magnify the complexities of mixture interpretation. The work will involve development of computer algorithms for likelihood ratios under competing hypotheses to explain the forensic evidence, taking into account genetic principles and appropriate models of laboratory processes. The resulting programs will be tested using simulations and real data.
In later stages of the project, the student will also help assess the potential for new technologies in forensic identification, by developing algorithms and computer code for robust evaluation of the evidence arising from such methods in advance of their use. The exact nature of the technologies to be assessed will depend on developments over the coming 18 months but they could include new DNA sequencing techniques for specific forensic applications, and/or the use of methylation signatures as evidence. Although less stable than DNA, methylation signatures can convey information not captured by the DNA, such as identifying the tissue type, and possibly the age, smoking status and other characteristics of the source individual.
The student will be co-supervised by Matt Greenhalgh, Cellmark's Director of Forensic Science. S/he will spend a substantial proportion of the PhD embedded in the R&D team at Cellmark, where s/he will receive training in specific techniques relevant to the research project, obtain experience of working with actual cases, as well as experience of working in a busy commercial-sector scientific laboratory.
Applicants should have either a formation in mathematics, statistics or computer science, together with a strong interest in genetics and in practical applications in forensic science, or a formation in genetics with a strong computational and/or statistical component.
Please contact Professor David Balding by email ([email protected]
) by Thursday March 22, 2012. Applications should include a CV and the names, positions, and e-mail addresses of two referees.