Improved Evolutionary Hybrids for Flexible Ligand Docking in AutoDock

Abstract

In this paper we evaluate the design of the hybrid EAs that are currently used to perform flexible ligand binding in the Autodock docking software. Hybrid evolutionary algorithms (E&) incorporate specialized operators that exploit domain-specific features to accelerate an EA'S search. We consider hybrid EAs that use an integrated local search operator to r e h e individuals within each iteration of the search. We evaIuate several factors that impact the efficacy of a hybrid EA, and we propose new hybrid EAs that provide more robust convergence to low-energ d d i n g configurations than the methods currently available in Autodock.

LDRD Final Report: Global Optimization for Engineering Science Problems

“…Furthermore, we evaluated strategies for adaptively initializing the step length in the local search to improve the efficiency and robustness of the optimization. Preliminary experiments with standard test problems suggest that these strategies can significantly improve the robustness of these methods, though our work with a drug docking application did not confirm these results [20].…”

“…We worked with Richard Belew (UCSD) to develop new techniques for hybridizing EAs on continuous domains [20,26]. Our model for hybridization is to refine a subset of the population of candidate solutions with a local search that is run for a bed duration.…”

“…Our research has focused on using direct search methods for local search. We performed comparison tests between different direct search methods to confirm that methods with better theoretical convergence guarantees led to better hybrid EAs [20]. Furthermore, we evaluated strategies for adaptively initializing the step length in the local search to improve the efficiency and robustness of the optimization.…”

“…We have successfully applied self-adaptive EAs, EPSAS, and hybrid EAs to this application [17,20,26,30]. The hybrid EAs are particularly effective at finding optimal solutions, and our hybrid EAs have found the best solutions for a testbed of six standard docking problems [20].…”

A Self-Adaptive Hybrid Genetic Algorithm for Color Clustering

“…Among the factors that affect this balance is the duration of local search [14], which is defined as the number of the consecutive local search iterations that is performed on a solution before terminating a local search procedure. This control parameter can be used to adapt the hybrid on-line to a specific problem.…”