Illustration of WAP scores.

A: Gene products are scored for their association x_i (number of interactions) to the top i DEGs, where i varies from 1 to the total number n of genes. Values of x_i are compared via edge-count probabilities P_i, and a gene product is scored with its best corrected attachment p-value: c min_i P_i. B: The profile of ranked c min P values is compared to profiles estimated after randomly assigning disease and control labels to samples (10⁴ permutations). This yields an estimated Proportion of False Positive (PFP) of WAPs at each rank j.