MULTIRESOLUTION SAMPLING METHODS FOR PROTEIN & PEPTIDE CONFORMATIONAL SPACE

点击翻译按钮获取中文摘要

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. This core project focuses on the development, validation, and application of hybrid models designed to explore aspects of protein and peptide folding, ligand docking and thermodynamics. The ability of the AMBER99/GBSA potential to recognize the native structure as being lowest in energy as well as its correlation with native-likeness was assessed on 150 proteins, a subset of the PDB200 benchmark set of nonhomologous structures representative of the PDB between 41 and 200 residues. For each protein, 14,000 decoys, previously generated with TASSER. were used. We selected a subset of 50 proteins for extended MD runs of 2 ns. In only 33% of the cases did the native trajectory snapshot have the lowest energy among decoys, and the average native-decoy energy gap was ¿22 kcal/mole. Moreover, the AMBER/GBSA energy does not generally exhibit a correlation with native-likeness as assessed by their RMSD from native, even when the relative weights of the AMBER/GBSA potential are optimized. Next, different versions of AMBER/GBSA force field will be tested and ways to improve the scoring abilities of the potential, based on decoy analysis, will be developed.by TASSER, were used. When the decoys were only locally minimized with AMBER/GBSA, nearly all of the native structures from the set are recognized as the lowest energy structures, with an average native-decoy energy gap of around 10% of the total native energy. The few failures involved proteins that were crystallized as a part of a complex. In a more difficult test, the decoys were subject to 100 ps Molecular Dynamics relaxation at 300¿ K followed by minimization. Now, the native structure is not the lowest energy one, and the AMBER/GBSA energy does not generally exhibit a correlation with native-likeness as assessed by the root-mean-square-deviation from native. In the coming year, the entire PDB200 set will be evaluated and an optimization of the AMBER/GBSA force field will be performed.