What's new in JKlustor, LibMCS
his workshop discusses the novel clustering technique, the maximum common substructure (MCS) based hierarchical clustering available in the JKlustor tool.
The concept of the MCS and its variants is introduced along with the analysis of the complexity of the MCS problem. Common approaches to finding the MCS of two structures are discussed and compared. The approach taken in ChemAxon's clustering tool, LibMCS, is elaborated, including the description of the datastructures and algorithms applied.
Problems emerging in MCS based clustering are studied; strategies and heuristics applied in LibMCS are outlined.
Among various potential application areas the analysis of VHTS screens is presented.
Future development of LibMCS and JKlustor is discussed, and the workshop is finished by an open discussion when the audience can suggest further features, different strategies, applications etc.