Figure 2.

Selected results from a de-novo motif discovery run. (A) Summary table of discovered motifs. (B) Sequence logos of order 0, 1 and 2 for one discovered motif. (C) Motif quality analysis and positional distribution. In the dataset-centered analysis (left) all input sequences are defined as positives. In the motif-centered analysis (right), only input sequences carrying a motif occurrence are positives. Their fraction is estimated using fdrtool (orange broken line on the upper right). The quality of motifs is quantified by average recall (AvRec), the blue area under the TP-to-FP-versus-recall curves. The curves for positive-to-negative ratios in the dataset of 1:1, 1:10 and 1:100 are plotted. Recall = TP/(TP + FN), where TP = true positives, FP = false positives, FN = false negatives. Positional distribution of the motif occurrences relative to the center of the sequences is shown on the bottom. (D) List of database motifs similar to discovered motif.