Fig. 1
From: AGeNNT: annotation of enzyme families by means of refined neighborhood networks
Excerpts of two rGNNs of IPR000312 (TrpD and homologs). a Using Cytoscape, all nodes with less than 150 sequences were deleted from the rGNN and the position of nodes was interactively altered for ease of interpretation. “Hubs” (circles) represent sequence clusters from the corresponding SSN and “spokes” (hexagons, labeled with Pfam-IDs) protein functions found within the respective GN. The diameter of the elements corresponds to the number of sequences. Clusters 6 and 37 possess distinct GNs that differ from the GN of clusters 65 and 97. Due to the parameters selected for SSN generation, these isofunctional sequences constitute two sequence clusters. Cluster 65 comprises sequences from different bacterial phyla; sequences of cluster 97 are mainly from Bacteroidetes. b This rGNN was created in the same manner as the one shown in (a). However, subspecies filtering was not applied. This is why this rGNN contains cluster 28, which is a highly specific GN of few proteobacterial species. Cluster 88 corresponds to cluster 65 shown in (a)