Algorithmic Cheminformatics

Summary

Cheminformatics is the application of algorithms, combinatorial approaches, and formal methods from Computer Science to problems in Chemistry. While being formally a very old research field, building the theoretical foundations for Cheminformatics seen from the perspective of state-of-the-art theoretical Computer Science is not at all established research. The second edition of the seminar on ``Algorithmic Cheminformatics'' brought together researchers working in Chemistry, Cheminformatics, and most importantly the relevant fields in Computer Science related to it. In contrast to the first Dagstuhl meeting in 2014, we specifically focused on the analysis of the behaviour of chemical systems in terms of reaction networks. This includes both networks inferred from experimental data, as well as networks implicitly specified by for example formal grammars. We integrated experts in concurrency theory, in particular using process calculi, Petri nets, and related formal approaches. State-of-the-art results in these fields are hardly used to infer qualitative and/or quantitative properties of chemical reaction systems, which are highly concurrent systems by nature. Most current modeling approaches in chemistry are either very abstract and aimed at formal algebraic properties of reaction networks, or use precise modeling on a very fine grained level such as the quantum mechanical one where computational costs prevent handling of more than a few molecules. In this seminar we therefore seeked to advance discrete modeling approaches for Systems Chemistry. In addition to bringing together the experts in the respective fields from Computer Science, we also invited wet-lab chemists in order to cross-fertilize the fields and generate mutually beneficial activities.