Computational chemical biology: identification of small molecular probes that discriminate between members of target protein families

Research in chemical biology focuses on the use of small molecules to study protein functions and distinguish between different targets and their functional properties. In this context, it would be helpful to better understand which currently available active compounds have a potential to differentiate between different targets belonging to a protein family. Such compounds might be utilized as a starting point for the development of small molecular probes that distinguish one or more targets from others within a family of interest. To address this question, we have designed a computational approach for data mining that involves the generation and quantitative assessment of selectivity profiles for compounds active against a protein family. Selectivity profiles were generated and represented in a consistent manner, and an intuitive weighting scheme was applied to account for the target differentiation potential of individual compounds. Based on a thorough analysis of public domain compound data, we have prioritized currently available active small molecules that displayed a tendency to differentiate between targets belonging to 15 different families. We have been particularly interested in identifying compounds having the highest general target differentiation potential within a family (rather than identifying compounds that are selective for one target over one or more others). These compounds might be utilized in target profiling and as starting points for further chemical exploration and probe generation. A compendium of prioritized active compounds with informative selectivity profiles is provided.