Compound stutter in D2S1338 and D12S391. (March 2019)

• Record Type: Journal Article
• Authors: Woerner, August E.; King, Jonathan L.; Budowle, Bruce
• Is Part Of: Forensic science international. Volume 39(2019)
• Page Start: 50
• Page End: 56
• Publication Date: 2019-03
• Subjects: Stutter -- Flanking variation -- Probabilistic genotyping -- Massively parallel sequencing; Forensic genetics -- Periodicals; Génétique légale -- Périodiques; Forensic genetics; Electronic journals; Periodicals; 614.1
• DOI: 10.1016/j.fsigen.2018.12.001
• Languages: English
• ISSNs: 1872-4973

Abstract

Highlights: Compound repeats can create compound stutter. Stutter products from a given motif are associated the uninterrupted stretch of that motif and other proximal nucleotides. Multivariate approaches permit joint predictions on both rates of stutter product formation. Abstract: Compound repeats provide a testing ground for evaluating hypotheses on the causes and consequences of stutter. As compound repeats have two repeating motifs, each motif may produce stutter variants; thus as the repeat within the locus is compound, so is the resultant stutter. Further, the rates of stutter formation between these motifs may not be independent. This lack of independence may complicate modeling strategies, thus contributing to the challenges of mixture interpretation that rely on nucleotide sequence. This study evaluates compound stutter in two STR loci: D2S1338 and D12S391. The effects of flanking variation, as well as possible interactions between the two different uninterrupted stretches (US) and their respective stutter variants are assessed. Multivariate multiple linear regression (MMLR) was used to show that, as with simple repeats, the rate of stutter product formation of a particular repeating motif is not solely a function of the US of that repeat. The nucleotides adjacent to the repeating motif also appear to influence the rate of stutter formation of that motif, with those nucleotides sometimes including the other motif. MMLR was used to estimate the size of these effects and to construct an example of a two-dimensional (thus, a compound) stutter prediction. This example may merit further investigation in the application of massively parallel sequencing data to mixture interpretation and probabilistic genotyping.