Skip to main content
. 2011 Sep 12;6(9):e24576. doi: 10.1371/journal.pone.0024576

Figure 4. Assessment of the tested motif discovery algorithms on target sets of gapped and un-gapped motifs.

Figure 4

(a) Summary of the success rates of the tested algorithms using different similarity cutoff. Using a CompareACE score cutoff of 0.65, 0.75 or 0.85, the success rates of each tool on the total target sets are shown. (b) Summary of the success rates of the tested algorithms on target sets of gapped and un-gapped motifs. All motifs were divided into two groups, gapped motifs or un-gapped motifs, based on the criteria of whether two or more continuous N characters were contained within the consensus sequence. Using a CompareACE score cutoff of 0.75, the success rates of each tool on the gapped and un-gapped motif target sets, were summarized. (c) Motifs successfully recovered only by Amadeus, AMD, MoAn or DME. Each successfully recovered motif is marked by a black-shaded box according to its CompareACE score (≥0.75). (d) Summary of the motifs specifically identified by AMD and SPACER. The numbers of gapped and un-gapped motifs specifically identified by AMD and SPACER are indicated. (e) Motif-level sensitivity and specificity of the nine tested tools on the target sets of gapped and un-gapped motifs.