A Kinematics and Evolutionary Approach for Modeling Molecular Conformations

摘要

This paper presents a novel kinematics and evolutionary approach called kinematics-based differential evolution (kDE) technique to model flexible biological molecules and to rapidly identify low-energy molecular conformations. The proposed kDE model consists of two modules: the pre-computation and the DE-loop. During pre-computation, a molecule is represented as a kinematics chain that adopts different conformations. At the DE-loop, a differential evolution algorithm is utilized as an optimization tool for directing the search towards low-energy molecular conformations. Computer implementation and results demonstrate the effectiveness of our approach in rapidly identifying a population of alternative, low-energy molecular conformations while attaining good convergence performance.