A vibrational entropy term for DNA docking with autodock. (June 2018)

• Record Type: Journal Article
• Authors: McElfresh, GW; Deligkaris, Christos
• Is Part Of: Computational biology and chemistry. Volume 74(2018)
• Page Start: 286
• Page End: 293
• Publication Date: 2018-06
• Subjects: Chemistry -- Data processing -- Periodicals; Biology -- Data processing -- Periodicals; Biochemistry -- Data processing; Biology -- Data processing; Molecular biology -- Data processing; Periodicals; Electronic journals; 542.85
• DOI: 10.1016/j.compbiolchem.2018.03.027
• Languages: English
• ISSNs: 1476-9271

Abstract

Graphical abstract: Including a vibrational entropy term improves ligand-DNA docking predictions with AutoDock. Highlights: Including a vibrational entropy term improves ligand-DNA AutoDock predictions. Vibrational entropy terms have been shown to improve the protein-ligand AutoDock predictions. The addition of a vibrational entropy term to AutoDock also improves DNA-ligand predictions. The length of the search process is important for DNA docking with AutoDock (with or without vibrational entropy terms). Abstract: DNA interacts with small molecules, from water to endogenous reactive oxygen and nitrogen species, environmental mutagens and carcinogens, and pharmaceutical anticancer molecules. Understanding and predicting the physical interactions of small molecules with DNA via docking is key not only for the comprehension of molecular-level events that lead to carcinogenesis and other diseases, but also for the rational design of drugs that target DNA. We recently validated AutoDock, a popular docking method that includes a physics-based scoring function and a Lamarckian Genetic Algorithm, for the prediction of small molecule geometries upon physical binding to DNA. In this work, we added a vibrational entropy term based on the docking frequency to the scoring function in order to improve the accuracy of the best (lowest) score geometry. We found that in four small molecule–DNA systems the inclusion of the vibrational entropy term decreased the root-mean-square-deviation from the experimental crystallographic structure. Including the entropy term also preserved the successful prediction of the binding geometry compared to the crystallographic structure for the rest of the small molecule–DNA systems. We also improved the method of creating clusters of docking geometries and emphasized the importance of the length of the search process for similar vibrational entropy terms.