NOCIt: A Computational Method To Infer the Number of Contributors to DNA Samples Analyzed by STR Genotyping

• NCJ Number: 249098
• Author(s): Harish Swaminathan; Catherine M. Orgicak; Muriel Medard; Desmond S. Lun
• Date Published: November 2014
• Length: 9 pages
• Grant Number(s): 2014-DN-BX-0001
• Agencies: NIJ-Sponsored
• Publication Type: Research (Applied/Empirical)
• Award: Computational methods for the interpretation of forensic DNA samples

Annotation

This article reports on the development and performance of a computational tool (NOCIt) that infers the number of contributions to a DNA sample.

Abstract

NOCIt correctly identified the number of contributors in 83 percent of 278 samples that contained between one and five contributors. Existing methods of identifying the number of contributors work on the number of peaks observed and/or allele frequencies. NOCIt works on single source calibration data that consists of known genotypes in computing the posterion probability (APP) for an unknown sample. This method takes into account signal peak heights, population allele frequencies, allele dropout, and stutter, a commonly occurring PCR artifact. NOCIt 's performance was tested with 278 experimental and 40 simulated DNA mixtures that consisted of one to five contributors. The number of contributors was correctly identified in 83 percent of the experimental samples and in 85 percent of the simulated mixtures. The accuracy of the best method currently being used was 72 percent for the experimental samples and 73 percent for the simulated mixtures. NOCIt calculated the APP for the true number of contributors to be at least 1 percent in 95 percent of the experimental samples and in all of the simulated mixtures. (Publisher abstract modified)