Prediction Of Functional Sites In 50,000 Protein Domains Using Dynamics Perturbation Analysis

Dynamics perturbation analysis (DPA) [1-3] finds regions in a protein structure where proteins are “ticklish,” i.e., where interactions cause a large change in protein dynamics. Such regions corresponded to native binding sites in a test set of hundreds of protein-ligand complexes [3]. We have further tested DPA for prediction of protein functional sites by applying an accelerated algorithm, Fast DPA [4], to over 50,000 structures of protein domains from the SCOP database [5]. We compared the predicted sites with known functional sites obtained from two sources: catalytic residues in the Catalytic Site Atlas; and putative binding sites identified by finding protein residues near small molecules in crystal structures. Combining predictions with information from sequence conservation and multiple sequence alignments reduced false positive rates. Overall, prediction of functional sites using DPA recapitulates much of the known information about functional sites in SCOP domains, and validates the use of DPA to predict functional sites in proteins. These results further suggest that functional sites in proteins tend to evolve at control points where interactions cause a large change in protein dynamics [1, 3]. 1. Ming D. and Wall M.E. Quantifying allosteric effects in proteins. 2005. Proteins 59:697-707. 2. Ming D. and Wall M.E. Allostery in a coarse-grained model of protein dynamics. 2005. Phys Rev Lett 95:198301. 3. Ming D. and Wall M.E. Interactions in native binding sites cause a large change in protein dynamics. 2006. J Mol Biol 358:213-223. 4. Ming D., Cohn J.D. and Wall M.E. Fast dynamics perturbation analysis for prediction of functional sites. 2008. BMC Struct Biol 8:5. 5. Murzin A.G., Brenner S.E., Hubbard T. and Chothia C. SCOP: a structural classification of proteins database for the investigation of sequences and structures. 1995. J Mol Biol 247:536-540.