SpaceMACS is a novel search method for similarity-driven virtual screening in large combinatorial fragment spaces. SpaceMACS identifies maximum common connected induced substructures between a query and all products of a fragment space, like commercial make-on-demand libraries. SpaceMACS heavily exploits the fragment structure of those spaces screening billions or even trillions of compounds within seconds to minutes on standard desktop computers. Runtimes depend on the number of results retrieved and the size of the fragment space representation. In general, they are mostly independent of the number of compounds represented. In order to explore fragment spaces, the SpaceMACS tool supports interactive command line sessions and processing of multiple queries avoiding the need to load the space for every single query.
The SpaceMACS tool is available to Linux,MacOS, and Windows as part of the NAOMI ChemBio Suite at https://software.zbh.uni-hamburg.de and is free for academic use and evaluation purposes.
To search with SpaceMACS you can access the KnowledgeSpace at https://www.biosolveit.de/infiniSee/?file=KnowledgeSpace_290tr_2019-05.space
People and references
For further reading on the new SpaceMACS tool please see:
Schmidt, R.; Klein, R.; Rarey, M. (2021) Maximum Common Substructure Searching in Combinatorial Make-on-Demand Compound Spaces . Journal of Chemical Information and Modeling