To allow for complex scoring methods including Hidden Markov Models, for cross-species conservation of motif occurrences, and for including thousands of features, we have separated the actual scoring and ranking of the 136K set from the enrichment calculation (Figure 1).