The Probabilistic Genotyping Software STRmix: Utility and Evidence for its Validity

Abstract

Forensic DNA interpretation is transitioning from manual interpretation based usually on binary decision-making toward computer-based systems that model the probability of the profile given different explanations for it, termed probabilistic genotyping (PG). Decision-making by laboratories to implement probability-based interpretation should be based on scientific principles for validity and information that supports its utility, such as criteria to support admissibility. The principles behind STRmix™ are outlined in this study and include standard mathematics and modeling of peak heights and variability in those heights. All PG methods generate a likelihood ratio (LR) and require the formulation of propositions. Principles underpinning formulations of propositions include the identification of reasonably assumed contributors. Substantial data have been produced that support precision, error rate, and reliability of PG, and in particular, STRmix™. A current issue is access to the code and quality processes used while coding. There are substantial data that describe the performance, strengths, and limitations of STRmix™, one of the available PG software.

A Review of Probabilistic Genotyping Systems: EuroForMix, DNAStatistX and STRmix™

“…Reference Ability of STRmix™ to deconvolute profiles and assign LRs that comport to manual interpretation and human expectation [15] Ability of STRmix™ to discriminate between donors and non-donors in database searches [190] Behaviour of STRmix™ to assign LRs when dealing with multiple replicates, different number of contributors, and assumed contributors [163] Sensitivity of LR produced by STRmix™ to different factors of uncertainty such as theta, relatedness of alternate DNA source and length of MCMC analysis [171] Tests to be performed when validating probabilistic genotyping, using STRmix™ as an example [112] Ability of individuals from different laboratories to standardise evaluations when using STRmix™ [33,53] Ability of STRmix™ to reliably use peak height information in very low intensity profiles [56,132,210] Ability of STRmix™ to discriminate between donors and non-donors in large-scale Hd true tests, or using importance sampling [59,60,190,200,21 2,213] Sensitivity of STRmix™ model parameters to laboratory factors [196,198] Ability of STRmix™ to utilise information from profiles produced under different laboratory conditions within a single analysis [155] Effect of mixture complexity, allele sharing and contributor proportions on the ability STRmix™ to distinguish contributors from non-contributors [54] The ability of STRmix™ to identify common DNA donors in mixed samples [25,159] The sensitivity of LRs produced in STRmix™ to the choice of the number of contributors [71,72,97] Ability to use STRmix™ to resolve major components of mixtures [72] Testing the assumption of additivity of peak heights in STRmix™ models [159,160] Performance of the degradation model within STRmix™ [214] The effect of relatedness of contributors to the STRmix™ analysis [203,215] Testing the calibration of LRs produced in STRmix™ [58] Validation overviews of STRmix™ [205,216] Comparison of STRmix™ ...…”

Developments in forensic DNA analysis

“…The software packages that implement probabilistic genotyping methods are highly complex, and developers have urged forensic laboratories to ensure their analysts have a good understanding of the concepts underlying the methods and that they remain involved in the interpretation of profiles and critical evaluation of the mixture analysis [36,49]. Concerns have been raised over variation in the output of probabilistic genotyping methods, some due to subjective decisions made by the user, some due to variability inherent in the methods [50,51].…”

Proposed Framework for Comparison of Continuous Probabilistic Genotyping Systems amongst Different Laboratories

“…Conditions 1 to 8 therefore provide for a minimum performance measure. Our condition 8 is a strenuous test because, as Bucketon et al [43] point out: "testing two low-level contributors with similar APHs (a 1:1 mixture) presents more of a challenge to the software than does a 1:20 mixture, as the genotype of the higher contributor has less uncertainty and helps to inform the genotype of the lower contributor". This would equally be the case at high APHs.…”

“…One way to achieve it is to amplify a dilution series of DNA such that there is a range of DNA template input amounts ranging from below the optimum to above the optimum. This is a general approach when assessing PG systems (e.g., [41,43,44]) and has been previously used to compare amongst them [29]. The LR will approach a maximum for H 1 true as DNA template amount increases and as w i → 0.…”