Identification evidence unraveled by strict proteomics rules toward forensic samples

Abstract Snake venom is a complex mixture of proteins and peptides secreted by venomous snakes from their poison glands. Although proteomics for snake venom composition, interspecific differences, and developmental evolution has been developed for a decade, current diagnosis or identification techniques of snake venom in clinical intoxication and forensic science applications are mainly dependent on morphological and immunoassay. It could be expected that the proteomics techniques directly offer great help. This work applied a bottom‐up proteomics method to identify proteins’ types and species attribution in suspected snake venom samples using ultrahigh‐performance liquid chromatography–quadrupole‐electrostatic field Orbitrap tandem mass spectrometric technique, and cytotoxicity assay was amended to provide a direct evidence of toxicity. Toward the suspicious samples seized in the security control, sample pretreatment (in‐sol and in‐gel digestion) and data acquisition (nontargeted and targeted screening) modes complemented and validated each other. We have implemented two consequent approaches in identifying the species source of proteins in the samples via the points of venom proteomics and strict forensic identification. First, we completed a workflow consisting of a proteomics database match toward an entire SWISS‐PROT (date 2018‐11‐22) database and a result‐directed specific taxonomy database. The latter was a helpful hint to compare master protein kinds and reveal the insufficiency of specific venom proteomics characterization rules. Second, we suggested strict rules for protein identification to meet the requirements of forensic science on improved identification correctness, that is, (1) peptide spectrum matches confidence, peptide confidence, and protein confidence were both high (with the false‐discovery ratio less than 1%); (2) the number of unique peptides was more than or equal to two in one protein, and (3) within unique peptides, which at least 75% of the ∆m / z of the matched y and b ions were less than 5 ppm. We identified these samples as cobra venom containing 10 highly abundant proteins (P00597, P82463, P60770, Q9YGI4, P62375, P49123, P80245, P60302, P01442, and P60304) from two snake venom protein families (acid phospholipase A2 and three‐finger toxins), and the most abundant proteins were cytotoxins.