Coarse-grained versus atomistic simulations: realistic interaction free energies for real proteins.

摘要: MOTIVATION: To assess whether two proteins will interact under physiological conditions, information on the interaction free energy is needed. Statistical learning techniques and docking methods for predicting protein-protein interactions cannot quantitatively estimate binding energies. Full atomistic molecular simulation do have this potential, but are completely unfeasible large-scale applications in terms of computational cost required. Here we investigate applying coarse-grained (CG) dynamics simulations a viable alternative complexes known structure. RESULTS: We calculate barrier with respect to bound state based using both full CG force field TCR-pMHC complex MP1-p14 scaffolding complex. find that barriers from similar accuracy as those ones, while achieving speedup >500-fold. also observe extensive sampling extremely important obtain accurate barriers, which only within reach models. Finally, show model preserves biological relevance interactions: (i) strong correlation between evolutionary likelihood mutations impact state; (ii) confirm dominant role interface core these interactions. Therefore, our results suggest can realistically be used prediction strength. Availability implementation: The python analysis framework data files available download at http://www.ibi.vu.nl/downloads/bioinformatics-2013-btt675.tgz.