Skip to main content
. 2010 Jan 22;6(1):e1000648. doi: 10.1371/journal.pcbi.1000648

Figure 1. The strategy workflow.

Figure 1

Circles and squares represent protein chains. Homologous chains share the same color. From each group of homologous chains, one is selected as representative and is shown as a square. (A) As the PDB has approximately 110,000 protein chains, identifying secondary targets directly is computationally intractable. (B) To reduce the number of chains, all chains are grouped by sequence homology into 12,646 clusters, and (C) a single representative chain is selected from each cluster. The set of all representative chains is called the PDB30. (D) SOIPPA is used to eliminate all protein chains in the PDB30 with active sites that are dissimilar to that of the primary target, TbREL1. Only 218 chains remain. (E) A new set of 645 protein-chain structures called the PDBr is created by taking the union of all those clusters whose representative PDB30 protein chains have active sites that are not dissimilar to that of TbREL1. (F) Redundant chains are ignored; compound 1 is docked into the remaining 87 protein chains. Chains are ranked by their corresponding docking scores.