Back to Presentations

Further Exploration of TrueAllele® Casework

S.A. Greenspoon, L. Schiermeier-Wood, and B.C. Jenkins, "Further Exploration of TrueAllele® Casework", Promega's Twenty Sixth International Symposium on Human Identification, Grapevine, TX, 13-Oct-2015.


Download Poster


The Virginia Department of Forensic Science (VDFS) began casework analysis with TrueAllele Casework (TA) in January of 2014. TA is used for analysis of complex mixtures comprised of up to four contributors on the most challenging of mixtures encountered by the laboratory. Since completion of validation, additional tests have been performed. These studies examined: 1) What effect does the use of the differential degradation feature have on the log(LR) of contributors to a differentially degraded mixture? 2) What happens to the TA analysis when either a greater or fewer number of contributors is solved for than are actually in the mixture? and 3) What happens to the log(LR) of contributors when the DNA sample is over-amplified/loaded?

The use of the differential degradation feature produced only a small (~1 log(LR) unit) change, if any, for contributors in differentially degraded mixtures. However, it may have affected how readily the sample was separated, resulting in fewer computer runs. This improvement may be due to a more accurate assessment of the mixture weights when differential degradation is taken into account.

When a greater number of contributors was hypothesized for TA analysis than was in the mixture, typically there was a small reduction, if any, effect on the log(LR) values generated. When a smaller number of contributors was hypothesized than was in the mixture, it often dramatically reduced the log(LR) of donors. This is consistent with how the TA modeling works; restricting the number of contributors also limits the potential genotype combinations that can explain the data, which may produce a reduction in the log(LR). Providing a greater number of potential contributors does not restrict the genotype combinations.

Without manual de-selection of excessive artifact peaks prior to TA analysis, the log(LR)s produced for contributor comparisons were typically decreased for over-amplified/loaded samples. The impact on log(LR) typically affected more minor contributors, rather than the predominate contributor. This result is consistent with how the TA process works; over-amplified/loaded samples exhibit excessive artifact which increases genotype uncertainty and thus can reduce the log(LR) values.

The results of the additional studies were consistent with the earlier work performed at VDFS and what has been reported in the literature for TA.