Text

theses

Biological evidence submitted to the forensic DNA laboratory contains cells from an unknown number of contributors in unknown proportions, resulting in profiles that are difficult to interpret.

Thus, recent efforts have focused on developing single-cell forensic DNA pipelines to deconvolve mixture signal by separating cells at the front end of processing. Single-cell signal, however, are often obfuscated by the presence of confounding signal such as false negative detection of alleles (i.e., drop-out); stutter, a polymerase chain reaction (PCR) artifact; and false positive detection of alleles (i.e., drop-in). Given the need to provide the weight-of-evidence against the accused, probabilistic characterization of the confounding single-cell artifacts is a necessity.

As such, 556 single-source, single-cell samples of known genotype were analyzed. The data were evaluated to determine if distributions associated with allele detection, stutter, and allelic drop-in were significantly different from those of bulk-processed samples. The results demonstrate that, in contrast to bulk-processed samples, allele detection is cell dependent. Like bulk-processed samples, stutter in the single-cell regime was found to be locus dependent; however, single-cell samples resulted in higher stutter ratios. As predicted, the frequency of allelic drop-in appeared consistent with that of bulk-processed samples. These findings suggest current state-of-the-art probabilistic systems are ill-equipped to evaluate single-cell evidence and new probabilistic constructs are required. The results of this study form the foundation from which these new inference systems may be developed.

Not only is probabilistic characterization of single-cell signal a necessity; practical implementation of a single-cell pipeline (i.e., that the cells can be effectively desorbed from common collection material such as cotton swabs) must be verified. Therefore, the second phase of this work focused on developing and accessing a protocol to desorb buccal cells from cotton-tipped applicators. To measure its efficacy, hemocytometry was used to determine the percent of cells recovered. The percent recovery of buccal cells appeared consistent with that of bulk-mixture extraction, demonstrating that a single-cell strategy is a viable alternative to the traditional forensic DNA pipeline.

ETD_9915

M.S.

Includes bibliographical references

doi:10.7282/t3-tq0z-7w76

ETD graduate

The author owns the copyright to this work.