Compounds can be associated by their shared molecular scaffolds, where a scaffold consists of one or more ring-systems joined by linkers.  These scaffold associations are meaningful and useful in organic chemistry, medicinal chemistry and chemical biology.  The importance of scaffolds is understood to derive both from (1) physical associations with compound properties including bioactivity, and (2) artifactual and other human-based associations such as synthesis/optimization strategies (analog series).  Cheminformatics methods and tools have existed to analyze and process scaffolds, and there is extensive literature relating to scaffold based methodology, but also problems, including lack of a rigorous definition of scaffold.  Given the strong consensus that scaffolds are important, and the prevalence of scaffold based approaches, the lack of standards for scaffold analysis in cheminformatics is notable.  We describe a set of scaffold analysis tools and methods developed by our group, available via open-source project UNM-biocomp-Hscaf, and examples of their use applied to the (1) CARLSBAD database, and (2) the NIH Molecular Libraries Program system BARD (BioAssay Research Database).  The scaffold analysis algorithm by Wilkins et al. was implemented and extended.  We show that scaffold associations can reveal patterns of bioactivity and promiscuity.  Also importantly, scaffold based patterns are inherently comprehensible to chemists, thereby facilitating hypothesis generation.  Thus, scaffold analysis is a powerful cheminformatics approach, a kind of chemical indexing, which can enable scientists to navigate biological space, and facilitate knowledge discovery in  realms such as chemical biology and drug discovery.

Molecular scaffolds are special and useful guides to discovery

246th National Meeting (2013)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.