PNNL

Abstract

Few tools exist to assess taxonomic origin of unknown/forensic biological samples using untargeted shotgun proteomics. Conventional proteomics approaches using protein database searching and identifying “unique” peptide markers makes assumptions about source organisms and suffers from signal erosion, respectively. To overcome these limitations, we describe and examine the performance of MARLOWE, an untargeted, statistics-based metaproteomics tool for application to sample types encountered in forensic evidence. Using publicly-available proteomics data from forensically-relevant samples, MARLOWE characterizes samples based on their protein profile and returns ranked organism lists of potential contributors and taxonomic scores based on few shared strong peptides between organisms. Overall, the correct characterization rate ranges between 44 and 100%, depending on the sample type and data acquisition parameters. MARLOWE demonstrates successful characterization to true contributors and close relatives, and provides sufficient specificity to distinguish within a bacterial superspecies group. Further, comparison of taxonomic score distributions revealed greater proteome similarity of primates than a bacterial superspecies group. MARLOWE demonstrates its ability to provide insight into potential taxonomic sources for a wide range of sample types in an entirely untargeted manner, which can find utility in forensic science and also broadly in bioanalytical applications that utilize metaproteomics approaches for source attribution.