Additional file 3 of A method for estimating coherence of molecular mechanisms in major human disease and traits

Additional file 3: Phenotype-specific coherence estimates. The PPI database (STRING, STRING filtered, Biogrid) are specified in the corresponding column names. For each phenotype category, the results are sorted alphabetically. In addition to the normalized coherence and the permutation p-values indicating significant difference of phenotype-specific coherence from random, modularity, untransformed slopes, and the corresponding lower and upper confidence interval bounds are shown. “NA” indicates the corresponding value cannot be estimated due to lack of sufficient number of genes annotated with PPIs.