Validation Study of the TrueAllele Automated Data Review System

• NCJ Number: 206523
• Journal: Journal of Forensic Sciences Volume: 49 Issue: 4 Dated: July 2004 Pages: 660-667
• Author(s): Kristy Kadash Ph.D.; Brian E. Kozlowski M.F.S; Lisa A. Biega M.S.; Barry W. Duceman Ph.D.
• Date Published: July 2004
• Length: 8 pages
• Publication Format: Article
• Publication Type: Test/Measurement; Report (Technical)
• Language: English
• Country: United States of America
• Note: This work has been presented orally at the following meetings: 28th Annual Meeting of the Northeast Association of Forensic Scientists, Atlantic City, NJ, November 2002; SWGDAM Annual Meeting at FBI Headquarters, Quantico, VA, January 2003; and the 55th Annual Meeting of the American Academy of Forensic Scientists, Chicago, IL, February 2003.

Annotation

This paper describes the materials, methods, and results of a validation study of the TrueAllele automated data review system, which is intended to address the shortage of skilled DNA data reviewers by automating steps in the review process, thus decreasing the amount of time required to analyze DNA profiles.

Abstract

The automated process begins with the creation of a datadisk that consists of the gel or capillary data; rule thresholds used for quality assessment; and user preferences, including short tandem repeat (STR) panel, choice of size standard nomenclature of extraction and amplification controls, and ladder preference. During Allele Call, TrueAllele identifies and quantifies peaks and then prioritizes the samples according to a set of user-defined rules. A table presented in this paper describes each rule and the optimized parameters used by the New York State Convicted Offender DNA Databank. TrueAllele prompts the user to accept, edit, or reject potentially problematic calls in Allele View without reviewing calls that are given a high-quality score. The validation of TrueAllele consisted of an extensive optimization phase followed by a large concordance phase. The materials and methods of these phases are described in this paper. During optimization, the rule settings were tailored to minimize the amount of high-quality data viewed by the user. In the concordance phase, a large dataset was typed in parallel with the ABI 3700 software Gene Scan and Genotyper (manual review) and TrueAllele (automated review) for comparison of allele calls and sample state assignment. Only 1 significant difference was found out of 2,048 samples in the concordance study. In this case, True Allele showed a spike in the profile that was interpreted as a DNA peak by the analyst in Genotyper. TrueAllele was designed to review poor data and eliminate the need for complete reanalysis technical review. This validation study found TrueAllele to be reliable in performing this work. Based on this validation study, TrueAllele is being used by the New York State Convicted Offender DNA Databank. 5 tables, 2 figures, and 6 references