Molecular modeling and simulation for proteins and polymer membranes
Christian, Bope Domilongo
Date of Issue2017
School of Biological Sciences
A computational framework was used to expand the applicable length and time scale of molecular dynamics simulations (MD) by developing a multiscale CG parameterization approach for biomolecules, and an investigation on the adsorption mechanism of an endocrine disrupting compound onto polymer membranes. For a multiscale model to parameterize CG-ENM force fields and structure-based model, we developed an improve fluctuation matching method, which is based on the relative entropy method, pioneered by Shell and colleagues. This method includes a non-negativity constraint and Newton Raphson’s algorithm for iteration. Furthermore, this framework incorporates pairwise force constant correlations, which play an important role in the study of protein dynamics. Similar to the well-established fluctuation matching proposed by Lyman and coworkers, our heterogeneous ENM parameterization using the relative entropy method with non-negativity constraints, including a pairwise force constant correlation which plays an important role in the study of protein dynamics. Furthermore, fluctuation matching based on relative entropy method guarantees global optimum, with fast convergence. For the adsorption of 17-α ethinylestradiol (EE2), we investigate the adsorption mechanism of EE2 from wastewater using MD simulations of monomers and polymer membranes level to validate experimentally observed results. Our findings from monomers level simulations with the small molecules EE2, and testosterone (polyether sulfone-EE2, polysulfone-EE2, polyvinylidene fluoride-EE2, polyether sulfone-testosterone) and their analogous structures as well as polymer membranes simulations with the small molecules EE2 and testosterone (polyether sulfone-EE2 polyvinylidene fluoride-EE2 and polyether sulfone-testosterone) shows that the enhancement of binding affinity between PES and EE2 from both simulation set ups (monomer level and polymer membranes level) obtained from binding free energy, mean-squared displacement, diffusion coefficients, and insight atomistic interaction are attributed mainly to the π-π interaction and hydrogen bond. Our simulation findings are in agreement with the experimental results. Furthermore, to verify prediction veracity of our computational framework, additional systems are considered such as polyamide 612 at monomer and polymer membrane level and polystyrene membrane. In total, the binding free energies of all the systems considered in this study are in the order as; PES-EE2 > PES-testosterone > PS-EE2> PA612-EE2 > PVDF-EE2 > PVDF-testosterone. To the best of our knowledge, these results suggest that the microporous PES hollow fiber membrane system is the most cost-effective approach that can effectively and efficiently remove EE2 at low concentrations.