"Development of an Analytical Approach for a Robust Probability Model for LCN DNA Profile Interpretation"

DESCRIPTION (provided by the applicant): Forensic DNA analysis is one of the success stories of the application of molecular biology to helping society. However, there are parameters to consider that impact the robustness and reliability of forensic DNA analyses. One critical parameter is the amount of template DNA used in the polymerase chain reaction (PCR). Some forensic DNA laboratories utilize various approaches to obtain short tandem repeat (STR) profiles from low copy number (LCN) DNA samples. One approach is increasing the PCR cycle number to enhance the sensitivity of detection. However, this approach has limitations: stochastic effects during the PCR are so exacerbated that, for STR analysis purposes, peak height imbalance, allele drop out, and increased stutter occur. Thus, there is uncertainty in determining the significance of a DNA result obtained by LCN typing. Probability models have been proposed to address this uncertainty. However, better probability models that also reduce bias are needed so that the forensic scientist can convey the significance and uncertainty of the data generated by LCN typing. I propose a two-fold approach: 1) develop a more robust system that is less refractory to stochastic effects; and 2) given the more robust system, generate empirical data to test an in-house developed probability model. To achieve step 1, additives, which reduce the paused extension of primer, stabilize the polymerase, or reduce instability of the template strand, will be evaluated to reduce the stochastic effects of PCR amplification. I hypothesize that amplification enhancers will improve the amplification of LCN DNA samples, specifically by increasing efficiency and sensitivity of the PCR. Such more robust data may be better suited for developing and testing a probability model for assessing the significance of LCN typing. Preliminary data illustrate that my approach already can improve the robustness of the amplification, specifically by decreasing stutter. I propose a one-year study to better define LCN typing using two different approaches. Each approach assesses improving the robustness of the amplification as it relates to LCN DNA samples. The second, and final, phase will be to apply the data to test a statistical model for LCN typing interpretation. The two specific aims for this proposal are: 1) To determine the role of amplification enhancers in the reduction of stochastic effects and the increased efficiency and sensitivity of the PCR with LCN DNA samples; and 2) To develop and test a statistical model for LCN typing interpretation. The empirical data generated during this study will play a critical role in my long term goal: the development of a probabilistic model to ascertain uncertainty or risk for genetic data generated under less than ideal DNA typing conditions. The impact is that some samples that traditionally yield too little DNA for typing may become suitable for routine analysis and/or a better assessment of uncertainty can be applied to casework evidence. ca/ncf