This is the Final Summary of a project whose overarching goal was to use the power of bioinformatics to harness the vast amount of proteomic mass spectrometry body fluid data to develop a predictive model with strong statistical confidence in the identification of menstrual blood, distinguishing it from venous blood; venous blood and menstrual blood mixture; and venous blood mixed with other body fluids.
Compared to the more commonly tested forensic body fluids that have easily identifiable abundant marker proteins due to the biological functions these proteins perform in their respective body fluids, menstrual blood is a mixture of the uterine endometrium, vaginal secretions, and blood, which is most abundant. A confirmatory test for the identification of menstrual blood has long been a goal of the forensic community, because such a test would achieve a predictive model with strong statistical confidence that can identify menstrual blood as distinct from other body fluids. The main objective of this application was to determine whether big data can be used to differentiate between menstrual blood and venous blood. This is particularly important when blood stains at a crime scene may be a female victim’s menstrual period or where a violent sexual assault with vaginal trauma may be claimed as consensual intercourse with a woman during menses. Sample selection and collection are described, along with protein extraction and analysis. Preliminary results from this study suggest that proteomic quantitation may be used to accurately distinguish menstrual from venous blood, regardless of coagulation status and time of collection. Based on model performance, clotted venous blood is modestly more distinguishable from menstrual blood than unclotted venous blood, but the difference is not statistically significant.