Tracking Reactions with the Reaction InChI (RInChI)

My research group studies reactions using experimental methods and computational techniques. The first approach gives us rich data, rather slowly. DFT methods are accurate but also rather slow. We use them both to get quantitative information about reactions and also to develop qualitative models to help people understand reactions: what picture can you draw on your fume-cupboard to tell you which catalyst to use? The more we learn about reactions, the more we learn about their complexity. To understand them, we need as much data on reactions as possible. The Reaction InChI (RInChI) is an identifier which enables us to make connections between reactions from different data sources, even when the number of reactions being considered is very large. I will report the results of preliminary tests on it using the SAVI database of more than a billion reactions, and outline the power of the RInChI, the limits and our plans for further extensions. We need more reaction data, but this is useful only if we can analyse it using the resources that we have available. The RInChI is an important tool for addressing this challenge.