A recursive algorithm for efficient combinatorial library docking

Abstract

Due to the rapid development of combinatorial chemistry and high throughput screening, a new virtual screening scenario emerged. While previously the focus was on analyzing large collections of compounds available to the medicinal chemist, nowadays the search space is defined in the form of large, possibly virtual, combinatorial libraries. In this article we describe how the structure of combinatorial libraries can be exploited to speed up docking predictions. Based on our incremental construction method implemented in the docking software FlexX we developed a recursive scheme to traverse the combinatorial library space efficiently. We applied our docking algorithm to three libraries with sizes from a few hundred up to 20 000 molecules. In all cases, we are able to show that similar results are achieved as in a sequential docking of the library molecules. The computing time, however, can be reduced by a factor of up to 30 resultingin an average time of about 5 s per library molecule.