Abstract:

One of the fundamental challenges in the application of molecular dynamics (MD) simulations to study biomolecular function is the presence of complex high-dimensional energy landscapes with high barriers, which prohibit adequate sampling of relevant configurations on computationally affordable timescales. Enhanced/path sampling methods such as metadynamics, adaptive bias force, and weighted ensemble simulations are implemented for overcoming these energetic and/or entropic bottlenecks. More specifically, we carried out well-tempered metadynamics in combination with time-lagged independent component analysis (tICA) to predict the relative significance of residues during the unfolding process of a protein and we also constructed a machine-learning based collective variable to stratify trajectories and calculate the kinetics using weighted ensemble simulations. The approaches are also extended to study the passive permeation of skin-oil oxidation products through the outermost layer of the skin, stratum corneum (SC), and important properties such as the free energy cost of solute partitioning, position-dependent diffusion constants and permeability coefficients are computed. Here, we highlighted the challenges in sampling an ordered lipid matrix like the SC membrane and also emphasized on the practical challenges in the implementation of contemporary approaches in the estimation of position-dependent diffusion coefficients. As part of this work, we also introduced the residence time analysis method, which provides a robust and computationally tractable framework derived from the principles of first-passage theory to evaluate the diffusivity profiles for solute translocation in lipid membranes. In our final study, we integrated the weighted ensemble technique with ab initio MD simulations to investigate the bioluminescence reaction of dinoflagellate luciferin and performed Non-adiabatic molecular dynamics (NAMD) simulations to probe the subsequent chemiexcitation process. In summary, this thesis demonstrates the application of enhanced sampling methods to understand various biophysical systems and to evaluate key thermodynamic and kinetic variables.