Random chemistry

Summary

This article discusses a new approach to the generation of high-diversity libraries of small organic molecules for drug discovery. The basis of the approach lies in an analysis of chemical reaction graphs, and the existence of a phase transition from subcritical to supracritical behavior as the diversity of substrates and candidate catalysts in the reaction system increases. This phase transition is related to the formation of ‘giant components’ in random graphs, and corresponds to the emergence of a giant connected web of catalyzed reactions in a reaction graph when a sufficient number of the reactions among the organic molecules find catalysts among a candidate set of catalysts. Supracritical reaction systems explode the molecular diversity present in a founder set of substrates. Within this high-diversity library, the presence of candidate molecules of interest is identified by assays such as displacement of a hormone from its receptor. Once the presence of such a product in the reaction system is detected, sib selection on repeated versions of the reaction system, using subsets of the founder substrates and of the candidate catalysts, winnows down to the critical substrates and catalysts which react to create the molecule of interest. As such, random chemistry appears to be a novel approach to aspects of organic synthesis and should prove useful in drug discovery. In addition, it may be an experimental avenue leading to the experimental formation of collectively autocatalytic sets of molecules.