Introducing SpaceGA:\nA Search Tool to Accelerate Large\nVirtual Screenings of Combinatorial Libraries

The growth of make-on-demand libraries in recent years\nhas provided\ncompletely new possibilities for virtual screening for discovering\nnew hit compounds with specific and favorable properties. However,\nsince these libraries now contain billions of compounds, screening\nthem using traditional methods such as molecular docking has become\nchallenging and requires substantial computational resources. Thus,\nto take real advantage of the new possibilities introduced by the\nmake-on-demand libraries, different methods have been proposed to\naccelerate the screening process and prioritize molecules for evaluation.\nHere, we introduce SpaceGA, a genetic algorithm that leverages the\nrapid similarity search tool SpaceLight (Bellmann, L.; Penner, P.;\nRarey, M. Topological similarity search in large combinatorial fragment\nspaces. <i>J. Chem. Inf. Model.</i> <b>2021</b>, <i>61</i>, 238–251). to constrain the optimization process\nto accessible compounds within desired combinatorial libraries. As\nshown herein, SpaceGA is able to efficiently identify molecules with\ndesired properties from trillions of synthesizable compounds by enumerating\nand evaluating only a small fraction of them. On this basis, SpaceGA\nrepresents a promising new tool for accelerating and simplifying virtual\nscreens of ultralarge combinatorial databases.