Research Article
Comparative Study of Evolutionary Algorithms for Protein-Ligand Docking Problem on the AutoDock
@INPROCEEDINGS{10.1007/978-3-030-32216-8_58, author={Zhuoran Liu and Changsheng Zhang and Qidong Zhao and Bin Zhang and Wenjuan Sun}, title={Comparative Study of Evolutionary Algorithms for Protein-Ligand Docking Problem on the AutoDock}, proceedings={Simulation Tools and Techniques. 11th International Conference, SIMUtools 2019, Chengdu, China, July 8--10, 2019, Proceedings}, proceedings_a={SIMUTOOLS}, year={2019}, month={10}, keywords={Evolutionary computation Swarm intelligence Protein-ligand docking Search function}, doi={10.1007/978-3-030-32216-8_58} }- Zhuoran Liu
Changsheng Zhang
Qidong Zhao
Bin Zhang
Wenjuan Sun
Year: 2019
Comparative Study of Evolutionary Algorithms for Protein-Ligand Docking Problem on the AutoDock
SIMUTOOLS
Springer
DOI: 10.1007/978-3-030-32216-8_58
Abstract
AutoDock is a widely used simulation platform for Protein-ligand docking which is a simulator to provide the field of computer-aided drug design (CADD) with conveniences. Protein-ligand docking establishes docking models and study interaction between the receptor and the ligand, as a part of the most important means in drug development. Protein-ligand docking problem is of great significance to design more effective and ideal drugs. The experiments are simulated on AutoDock with six weighted algorithms such as Lamarckian genetic algorithm, a genetic algorithm with crossover elitist preservation, artificial bee colony algorithm, ABCDEbased hybrid algorithm, fireworks algorithm, and monarch butterfly optimization. The diversity of search function constructed by different evolutionary algorithms for separate receptors and ligands is applied and analyzed. Performances of distinct search functions are given according to convergence speed, energy value, hypothesis test and so on. This can be of great benefit to future protein-ligand docking progress. Based on the work, appearances are found that performances of the same algorithm vary with different problems. No universal algorithms are having the best performance for diverse problems. Therefore, it is important how to choose an appropriate approach according to characteristics of problems.
