A sensitivity analysis to determine the robustness of STRmix™ with respect to laboratory calibration

Hannah Kelly; Jo-Anne Bright; Maarten Kruijver; Stuart Cooper; Duncan Taylor; Kyle Duke; Melissa Strong; Vickie Beamer; Christina Buettner; John Buckleton

doi:10.1016/j.fsigen.2018.04.009

A sensitivity analysis to determine the robustness of STRmix™ with respect to laboratory calibration

Forensic Sci Int Genet. 2018 Jul:35:113-122. doi: 10.1016/j.fsigen.2018.04.009. Epub 2018 Apr 27.

Authors

Affiliations

¹ Institute of Environmental Science and Research Limited, Private Bag 92021, Auckland, 1142, New Zealand. Electronic address: Hannah.kelly@esr.cri.nz.
² Institute of Environmental Science and Research Limited, Private Bag 92021, Auckland, 1142, New Zealand.
³ Institute of Environmental Science and Research Limited, Private Bag 92021, Auckland, 1142, New Zealand; Department of Mathematics, VU University, Amsterdam, Netherlands.
⁴ Forensic Science South Australia, 21 Divett Place, Adelaide, SA 5000, Australia; School of Biological Sciences, Flinders University, GPO Box 2100, Adelaide, SA 5001, Australia.
⁵ California Department of Justice Bureau of Forensic Services, United States.
⁶ San Diego Police Department Crime Laboratory, CA, United States.
⁷ Scottsdale Police Department Crime Laboratory, AZ, United States.
⁸ Wyoming State Crime Laboratory, United States.
⁹ Institute of Environmental Science and Research Limited, Private Bag 92021, Auckland, 1142, New Zealand; University of Washington, Department of Biostatistics, Seattle, WA 98195, United States.

PMID: 29727813
DOI: 10.1016/j.fsigen.2018.04.009

Abstract

STRmix™ uses several laboratory specific parameters to calibrate the stochastic model for peak heights. These are modelled on empirical observations specific to the instruments and protocol used in the analysis. The extent to which these parameters can be borrowed from laboratories with similar technology and protocols without affecting the accuracy of the system is investigated using a sensitivity analysis. Parameters are first calibrated to a publicly available dataset, after which a large number of likelihood ratios are computed for true contributors and non-contributors using both the calibrated parameters and several borrowed parameters. Differences in the LR caused by using different sets of parameter values are found to be negligible.

Keywords: Interpretation; Likelihood ratio; Mixtures; PCAST report; Probabilistic genotyping; Validation.

Publication types

Research Support, U.S. Gov't, Non-P.H.S.

MeSH terms

Alleles
DNA Fingerprinting
Forensic Genetics
Genotyping Techniques
Humans
Microsatellite Repeats*
Models, Statistical*
Probability*
Software*