Figure 1

Assessing the quality of the automatically generated clusters. This figure shows the automatically generated family-level clusters for the unclassified domains in the SCOP 1.69 "d" class (i.e. the alpha and beta proteins (a+b)). This figure also shows the representative domain structures for each cluster. A connected graph corresponds to an automatically detected MCL cluster. The ellipses indicate the novel families in SCOP 1.69. The MCL clusters are assigned a family-level label based on the most common family-level label in the cluster. Within a cluster, the nodes with the same color indicate that all these nodes have the same family-level label. To keep this figure simple, only clusters with more than four domains are shown. There are an additional of 79 clusters that matched the SCOP family label, and of these 30 clusters correspond to new families in SCOP 1.69. This figure was generated using BioLayout [31] and PyMol [32].