FIMO scans a sequence order database for personal coincidences to each of the motifs you ensure (sample output for list of motifs and sequences).
The name FIMO implies ‘Find Individual Motif Occurrences’. The pattern shows a database of sequence orders for occurrences of definite motifs, describing each motif independently.
Every user can determine the statistical threshold (p-value) for the list of motifs and whether FIMO copies just the appropriate sequence orders or their reverse complements, as well.
The pattern utilizes a dynamic programming algorithm to transform log-odds scores into p-values, assessing a zero-order background pattern.
The p-values for each motif occurrence are transformed to q-values following the method of Benjamini and Hochberg (“q-value” is determined as the minimal false discovery level at which a given motif occurrence is deemed significant). The –qv-thresh option directs the program to use q-values rather than p-values for the threshold.
If a motif has the strand feature set to +/- (rather than +), then FIMO will search both strands for occurrences.
The parameter –max-stored-scores sets the maximum number of motif occurrences that will be retained in memory. It defaults to 100,000. If the number of matches found reaches the maximum value allowed, FIMO will discard 50% of the least significant matches, and new matches falling below the importance rate of the retained featuring will also be discarded.