Principal components analysis (PCA). A, PCA based on 68 masses selected following Benjamini-Hoshberg FDR correction (p<2.5E-5), Bartlett’s test for sphericity (p<0.05), and a KMO greater than 0.81. Each point represents a patient profile, colored by disease state (black circles, control subjects; orange diamonds, PC patients). B, Scatter plots of the p-values (log10) for each of the 68 masses based on t-test comparisons between clinical variables (see legend). See Additional file 2 for PCA parameters and p-values.