Analyzing Protein Folding by High-throughput Simulations

Abstract

Molecular Dynamics allows investigating the dynamical properties of biomolecules. Protein folding simulations are computationally challenging when simulating all involved (solvent) atoms considering timescales of ms or slower. Native structure based models (SBM,‘Go-models') reduce computational complexity and have been proven to be a robust and efficient way for exploring the protein folding process (Schug and Onuchic 2010; Thirumalai, O'Brien et al. 2010). They are based on energy landscape theory and the principle of minimal frustration. Using this framework, we simulate protein folding for a large set (∼ 200) of non-homologous monomeric proteins sized from 50-150 amino acids in coarse-grained simulations, representing each amino acid by a single bead. A fully automatized workflow implemented with the help of eSBM Tools (Lutz et al.) guides these simulations. From the simulations, we extract typical folding properties like phi-values, folding free energy landscape and transition state ensembles. We repeat the simulations for a variant SBM with flavored contact strengths pending on amino acids composition. The resulting database estimates the robustness of folding parameters, quantifies the folding behavior, compares the behavior to existing experimental data and can serve as a baseline for comparison to future experiments or simulations of protein folding.Schug, A. and J. N. Onuchic (2010). “From protein folding to protein function and biomolecular binding by energy landscape theory.” Curr Opin Pharmacol 10(6): 709-714.Thirumalai, D., E. P. O'Brien, et al. (2010). “Theoretical perspectives on protein folding.” Annual review of biophysics 39: 159-183.Lutz, B., et al. (2012), in Proceedings of the 11th International Conference on Modeling and Applied Simulation, Wien, 2012, edited by M. Affenzeller, et al., pp. 237.