Target-Focused libraries: Rapid selection of potential PDE inhibitors from multi-million compounds' repositories
Reference Space was defined by collecting known PDE inhibitors ("seeds") from available literature and databases, and its property space was determined by applying the Lipinski and Veber parameters using Instant JChem. The searchable Chemical Space was composed by the recent edition of the 10 major vendor databases (cca. 10 M compounds). The ligand-based 2D similarity search (Instant JChem) was performed by setting a similarity treshhold (0.65-0.7 Tanimoto). The resulting set was shrunk to match the "reference property space" applying the appropriate physico-chemical parameter ranges. The filtered focused library was further analyzed to reach a reasonable diversity („seed” compound representation/ chemotype distribution) using Instant JChem and LibMCS and the process was completed by visual inspection. In the presentation we discuss our experience in the focused library generation and share the lessons we learnt during the application of the ChemAxon software family. This work was partially supported by the National Development Agency (NKÜ) grant: #KMOP 1.1.1-07/1-2008-0029)  Decornez H, Gulyás-Forró A, Papp A, Szabó M, Sármay G, Hajdú I, Cseh S, Dormán G, Kitchen DB Design, selection, and evaluation of a general kinase-focused library. ChemMedChem. 4(8), 1273-8 (2009).  A. Tovar, H. Eckert, J. Bajorath, Comparison of 2D Fingerprint Methods for Multiple-Template Similarity Searching on Compound Activity Classes of Increasing Structural Diversity, ChemMedChem, 2, 208-217 (2007).