Using these parameters, TADA-Denovo calculates the Bayes factors of all input genes. Next, it computes the p value for each gene by generating random mutational data, based on each gene’s specified mutation rate, to obtain a null distribution of Bayes factors. We used 1,000 samplings of de novo variants in each gene to determine null distributions. Finally, TADA-Denovo calculates a false discovery rate (FDR) q-value for each gene using a Bayesian “direct posterior approach.” A low q-value represents strong evidence for TD association.
