Forensic mixture interpretation is an arduous task that requires experts in multifarious fields to create data analysis and inference solutions that are fit for purpose. The difficulty in translating these technologies to operations has been effective and transparent model-building, testing and validation of these (often) black-box systems. To fill this gap we maintain the PROVEDIt Database. A fully open, free database of over 27,000 forensically relevant DNA mixtures.
We develop systems that enable the determination of a fullLR, which is a complete and overall evaluation of DNA data without application of signal thresholds, high-pass filters or subjective assessments of numbers of contributors. This includes the development of a software system, ValiDNA, engineered to work with most relevant propositions. The end goal is to provide full-weights-of-evidence using well characterized high fidelity data. We leverage our team’s track record in analytical chemistry and computer science and partner with experts in probability and statistics to build it.
We are exploring various means to apply single-cell analysis in the forensic sciences. This includes developing and coupling laboratory and interpretative approaches. The end goal is to produce a single-cell pipeline – from extraction to report – founded on sound analytical and probabilistic principles.
This work is supported by NIJ2011-DN-BX-K558, NIJ2012-DN-BX-K050, NIJ2014-DN-BX-K026, NIJ2018-DN-BX-K0185 and ARO RIF W911NF-14-C-0096 awarded by the National Institute of Justice, Office of Justice Programs, U.S. Department of Justice and the Department of Defense, Army Research Office, Rapid Innovation Fund, respectively.