Diagnostic Peptide Discovery: Prioritization of Pathogen Diagnostic Markers Using Multiple Features
The figure shows a number of enrichment plots obtained under different prioritization scenarios. In all plots: the x axis contains the prioritized proteome (top ranking proteins at the origin); the y axis displays the fraction of known validated antigens recovered in the top x proteins; the blue dashed line displays an hypothetical enrichment plot with an AUC = 0.5 (expected by chance), while the black solid line represents the actual enrichment obtained in each prioritization. From the top-left: comparison of different prioritization strategies (ordered by decreasing AUC values): 1) our composite method, 2–9) a number of prioritizations using a single criteria in each case: 2) Codon Usage bias (CAI), 3) Internal repeats, 4) Proteomic evidence of expression, 5) natively unstructured regions, 6) antigenicity (Bepipred), 7) surface localization (GPI), 8) O-Glycosylation, 9) antigenicity (EMBOSS antigenic). p-value, p-value (p-values based on a random permutation test, n = 10,000).