Reproducibility of complex proteomic signatures in detecting early stage ovarian cancer has proven to be a significant challenge in the field of biomarker discovery. Although early reports of proteomic profiling demonstrated near 100% sensitivity and specificity in discriminating ovarian cancer from non-cancer, the ability to reproduce these results across datasets has been difficult.