Abstract
The increased interest in the use of Massively Parallel Sequencing (MPS) technologies to type traditional autosomal STR markers raises multiple questions regarding interpretation of the results via probabilistic genotyping. To begin to address some of those questions, we examined the effects of using differing degrees of sequence information, pre-filtering, and data modeling to interpret complex MPS-STR mixtures in a probabilistic genotyping software. Sixty ForenSeq typing results for mixtures of from two to four contributors were: 1) represented using three separate formats that captured different degrees of sequence information, and 2) were analyzed using three different filtering approaches prior to probabilistic interpretation. All mixtures for the different format and filtering variants were subsequently interpreted with respect to ten reference profiles, using both qualitative (LRmix) and quantitative (EuroForMix) models to calculate the likelihood ratio (LR). The LR results indicated moderate information gain when the STR nomenclature was based upon the longest uninterrupted stretch (LUS) compared with conventional capillary electrophoresis repeat units (RU), whereas additional gains were very small when the complete sequence information was utilised. Use of a static analytical threshold for data pre-filtering improved LRs compared to a dynamic (percentage-based) threshold, as the static threshold prevented excessive filtering of alleles originating from minor contributors. For interpretations performed using a quantitative model, a small improvement in performance was observed if a stutter model was employed instead of using stutter thresholds to pre-filter the data, whereas – as expected – performance worsened considerably under the qualitative model when stutter was not pre-filtered. Given the empirical and theoretical findings in this study we discuss the value of utilizing sequence-level information and potential paths forward to increase information gain using MPS systems.
Highlights
Over the past five years, Massively Parallel Sequencing (MPS) has become more affordable and its use for sequencing DNA extracted from biological samples has been widely accepted by many forensic laboratories
Formats (Fig. 1A) and the longest uninterrupted stretch (LUS) with LUS+ formats (Fig. 1B) when the dynamic filter was used and the quantitative model was employed for interpretation
The examination of empirical ForenSeq assay data in this study indicated an approximately 1.2 increase in the log10LR values when using some sequence information as compared to no sequence information for true contributors for all the filter methods when the quantitative model was used for interpretation
Summary
Over the past five years, Massively Parallel Sequencing (MPS) has become more affordable and its use for sequencing DNA extracted from biological samples has been widely accepted by many forensic laboratories. Probabilistic genotyping software packages that are based upon a quantitative model, such as EuroForMix [1] and others (e.g., DNAmixtures [2], STRmixTM [3], TrueAllele© [4]; see Coble et al [5] for a review) were mainly designed for CE based data. Bleka et al [7] showed that sequence coverage information could be utilised in a quantitative model for autosomal SNP data developed via MPS, even though the model itself was designed for CE data. Hwa et al (2019) [8] performed mixture interpretation for both autosomal SNP and STR data generated using the ForenSeqTM DNA Signature Prep Kit, and demonstrated that EuroForMix could effectively use the sequence coverage depth to assign minor contributors, even for degraded samples
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
