Map Charts Chemical Space

A map covering the entire universe of possible small-molecule structures could, in the spirit of the Starship Enterprise, help researchers boldly go where none have gone before to discover new drugs and new materials. An algorithm to generate such maps has been created by a team led by Weitao Yang and David N. Beratan of Duke University (J. Am. Chem. Soc., DOI: 10.1021/ja401184g). The algorithm is capable of churning out virtual collections of molecules. Eliminating unstable and nonsynthesizable structures from the collections leads to compound maps that scientists can scout for drug candidates and advanced materials to synthesize. This isn’t the first effort to map chemical space. For instance, Jean-Louis Reymond of the University of Bern, in Switzerland, and coworkers developed a database of 166 billion small molecules last year (J. Chem. Inf. Model., DOI:10.1021/ci300415d). But the effort by Beratan and coworkers is the first attempt to map the entirety of chemical space, which includes an estimated 10⁶⁰ molecules.