In 2004, David R. Liu and coworkers at Harvard University developed a DNA-based system for discovering new chemical reactions, but the conditions required for DNA base-pairing restricted its utility. Now, they have eliminated the need for base-pairing in their second-generation platform, which works in organic solvents and at high temperatures, conditions that are hostile to base-pairing (J. Am. Chem. Soc., DOI: 10.1021/ja074155j). The new setup uses a redesigned DNA architecture in which the sequence of a single DNA strand serves as readable bar code for the identity of two reactants, A and B (shown); the first system encoded each reactant on separate strands. Hundreds of different pairs of reactants are covalently attached to encoding strands, but in each case, one reactant is connected by a biotin-bearing disulfide linker. The system enables the scientists to efficiently discriminate between reactant pairs that form a covalent bond (red) under a given set of conditions and those that don't. The team uncovered what they hope is the first of many new reactions-a mild coupling between indoles and some alkenes.