出版社：Academy & Industry Research Collaboration Center (AIRCC)
摘要：The screening of chemical libraries is an important step in the drug discovery process. Theexisting chemical libraries contain up to millions of compounds. As the screening at such scaleis expensive, the virtual screening is often utilized. There exist several variants of virtualscreening and ligand-based virtual screening is one of them. It utilizes the similarity of screenedchemical compounds to known compounds. Besides the employed similarity measure, anotheraspect greatly influencing the performance of ligand-based virtual screening is the chosenchemical compound representation. In this paper, we introduce a fragment-basedrepresentation of chemical compounds. Our representation utilizes fragments to represent acompound where each fragment is represented by its physico-chemical descriptors. Therepresentation is highly parametrizable, especially in the area of physico-chemical descriptorsselection and application. In order to test the performance of our method, we utilized an existingframework for virtual screening benchmarking. The results show that our method is comparableto the best existing approaches and on some data sets it outperforms them.
