Fig 1. The workflow of iFORM.

Overview of the iFORM workflow. (1) Assemble input data. The results may be improved by restricting the input to high-confidence sequences. Some programs achieve improved performance using phylogenetic conservation information from orthologous sequences or information about protein DNA-binding domains. (2) Choose several motif discovery programs for the analysis. (3) Test the statistical significance of the resulting motifs. Use control calculations to estimate the empirical distribution of scores produced by each program on random data. (4) For each identified TFBS, a combined p-value obtained from these five methods was calculated using Fisher's combined probability test. (5) FDR correction of the combined p-values obtained from Fisher’s method.