Written by Michael Coble, PhD, University of North Texas Health Science's Center for Human Identification and Jo-Anne Bright, PhD, Institute of Environmental Science and Research Limited (ESR)

One of the greatest challenges for the forensic DNA analyst is the interpretation of complex DNA mixtures. These include mixtures with multiple contributors (two or more) and/or low-level contributors where alleles may be missing from the profile (dropout). Exacerbating this problem is the type of evidence being submitted to the laboratory for testing with a movement away from high quality/high quantity samples like bloodstains to lower quality/lower quantity samples such as swabs from handguns and door knobs. Recent improvements in STR kit chemistry and Capillary Electrophoresis sensitivity have also increased the complexity of profiles recovered from exhibits and subsequent mixture interpretation. Finally, an increase in the number of core CODIS loci from 13 to 20 has had an impact on the time it takes to interpret a mixture.

Probabilistic Genotyping software can assist the analyst to interpret complex mixtures compared to manual methods used historically in forensic DNA testing. According to SWGDAM, probabilistic software uses biological modeling, statistical theory, computer algorithms, and probability distributions to calculate likelihood ratios (LRs) and/or infer genotypes for the DNA typing results of forensic samples. Over one-half of laboratories in the U.S. have now adopted probabilistic methods of interpretation for complex DNA mixtures.

The statistical output of probabilistic software is a LR. Although the LR was recommended as an appropriate method to calculate a statistic for mixtures by the DNA Advisory Board in 2000, until recently most laboratories in the U.S. reported either the Combined Probability of Inclusions (CPI) or the modified Random Match Probability.

As more and more laboratories brought probabilistic genotyping software online, we recognized the need to provide basic training in LRs to analysts, the judiciary and other interested parties. The ISHI meeting has historically dedicated time before and after the main symposium for workshops and training for the community. We first gave a half-day workshop on LRs at the 27th ISHI in Minneapolis in 2016, and then provided a full-day workshop at the 28th ISHI in Seattle in 2017. The workshops were well attended and we received a great deal of positive feedback – including a constant request for more worked examples (who knew LRs were so popular :>).

We approached the publisher CRC Press with an idea to write a book that was focused on explaining how to assign and report an LR and set propositions) for both simple and complex scenarios to scientists, lawyers, and judges who may have very little experience with LRs or statistics in general. We include topics such as assigning LRs considering relatives (for scenarios such as “it wasn’t me, it was my brother”) and provide step by step instructions demonstrating how LRs are assigned with examples of binary, semi-continuous and fully continuous probabilistic methods.

The book, Forensic DNA Profiling: A Practical Guide to Assigning Likelihood Ratios, was published in December 2019 and is available through the CRC Press website and other online book sellers such as Amazon and Barnes and Noble. We would like to thank all of the forensic scientists and lawyers that have encouraged us by attending those workshops, and of course ISHI for their continued dedication to provide training and education to the forensic DNA community.

Editor's Note: Learn how to argue the case for DNA evidence based on probabilistic genotyping at the workshop that Mike and Jo are leading at ISHI 31. This 3-hour workshop is scheduled for Sunday, September 13. The cost is only $100. Find the details and register on the ISHI website.