Validation Support

The forensic laboratory is a complex system where each step influences outcomes of subsequent steps

We will provide,

  • training and workshops that describe the theories, models, methods and tools used to validate the entire system in a cost- and time-efficient manner
  • support to optimize the laboratory conditions such as the PCR cycle parameters, PCR reaction volume, injection conditions for capillary electrophoresis and analytical thresholds
  • reports that demonstrate the laboratory conditions are fit-for-purpose and provide the highest allele information content for the technology employed. Reports include demonstrations of maximal signal:noise resolution, minimal false noise detections, minimal allele drop-out rates, minimal Pr(LR>1) for millions or billions of non-contributors and the most informative LR for true contributors
  • Contact Catherine Grgicak at c.grgicak@rutgers.edu to inquire
Schematic of the Forensic Laboratory Pipeline and its validation using ValiDNA – a software tool that optimizes the entire genotyping pipeline and demonstrates the validity of the process for forensic purposes. All data are generated using a PCR-based laboratory system and analyzed using a peak detection software of choice. The calibration data are garnered from single-source profiles of known genotype analyzed without the application of an analytical threshold. Well-characterized artifacts, such as pull-up and minus A, are filtered in an automated fashion using CleanIt and user-defined criteria. ReSOLVIt synthesizes data for multifarious scenarios, and the scenario resulting in the highest allele information content per EPG is set as the optimal laboratory condition. We confirm the laboratory conditions are optimized by evaluating the impact on downstream inference. We validate the optimized laboratory system by confirming good signal:noise resolution, low allele dropout rates, low noise detection rates, informative LRs and reasonable number of contributors (NoC) assessments. LRs and probabilities on the NOC are computed using state of the art probabilistic software modules engineered to evaluate the entire signal (i.e., NOCIt and CEESIt). NOCIt informs the analyst as to the NOC assumption(s) that ought to be considered when computing the summary statistics associated with evidence strength. CEESIt computes the LR distribution, the LR, and the probability that a randomly generated genotype would result in a LR>1.
ValiDNA report demonstrating optimal laboratory conditions improves inference. Parallel plots of Pr⁡(LR>1) for true non-contributors for experimental 1-, 2- and 3-person samples (left) injected for 5 s on the 3130 Genetic Analyzer coupled with an analytical threshold (AT) of 50 RFU and (right) injected for 20 s on the 3130 platform coupled with an optimized AT of 15 and 20 for the blue/green and yellow/red channels, respectively.  All samples were amplified with the Identifiler® Plus set of loci using 29 PCR cycles.