NCJ Number
228339
Journal
Forensic Science International: Genetics Volume: 3 Issue: 4 Dated: September 2009 Pages: 222-226
Date Published
September 2009
Length
5 pages
Annotation
This paper presents a statistical model for estimating the per locus and overall probability of allelic drop-out in forensic cases in which the amount of DNA is sparse at the crime scene, using the results of all STR loci in the case sample as reference.
Abstract
The authors demonstrate a simple and applicable way of assessing the drop-out probability of STR alleles in forensic genetics. The drop-out probabilities computed with the proposed model concur with the prior knowledge of the drop-out behavior varying with the observed peak heights. The methodology of logistic regression is thus appropriate for assessing drop-out probability. Future work will test the model on a larger dataset that includes more alleles. With a larger dataset, it may also be possible to test whether alleles or fragment length has a significant effect on the drop-out probability as the individual specific effect decreases with the number of different profiles. The authors advise that the drop-out probabilities may vary among laboratories, machinery within the same laboratory, and typing kits used for profiling. This is due to differences in the ability to amplify the DNA in the PCR and in the potential to measure the light intensities for the electropherogram. Consequently, before applying the proposed methodology in the likelihood ratio for evidence calculation, the laboratory should perform experiments with known profiles in order to estimate the parameters in the logistic regression model. The analysis was based on 175 controlled experiments conducted at the Section of Forensic Genetics, Department of Forensic Medicine, Faculty of Health Science, University of Copenhagen (Denmark). The experiments consisted of pairwise mixtures of four profiles and samples with only one contributor diluted in water. Descriptions of material and methods pertain to data and the logistic regression model. 5 tables, 1 figure, 8 references, and appended examples