Fig. 5From: A fast topological analysis algorithm for large-scale similarity evaluations of ligands and binding pocketsDissimilarity matrices for the Kahraman benchmark [12]. The colour code represented by the legend shows the dissimilarity score value. a Pairwise dissimilarity matrix for binding pockets using complete amino acid structures, b pairwise dissimilarity matrix for binding pockets using atoms within a 6.5 Å distance from the bound ligand, c pairwise dissimilarity matrix for ligand structures.Back to article page