Generic Strategies for Chemical Space Exploration

By Jakob Lykke Andersen, Christoph Flamm, Daniel Merkle, Peter F. Stadler. 2013

47 citations recorded by the source corpus.

The chemical universe of molecules reachable from a set of start compounds by iterative application of a finite number of reactions is usually so vast, that sophisticated and efficient exploration strategies are required to cope with the combinatorial complexity. A stringent analysis of (bio)chemical reaction networks, as approximations of these complex chemical spaces, forms the foundation for the understanding of functional relations in Chemistry and Biology. Graphs and graph rewriting are natural models for molecules and reactions. Borrowing the idea of partial evaluation from functional programming, we introduce partial applications of rewrite rules. A framework for the specification of exploration strategies in graph-rewriting systems is presented. Using key examples of complex reaction networks from carbohydrate chemistry we demonstrate the feasibility of this high-level strategy framework. While being designed for chemical applications, the framework can also be used to emulate higher-level transformation models such as illustrated in a small puzzle game.

Subfields: Computational Chemistry & Drug Discovery

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Chemistry & Drug Discovery

Citations in the corpus

Planning chemical syntheses with deep neural networks and symbolic AI (2017, methodology): “Actions are then transformations (rules/productions on graphs[50]) applied to one of the molecules m in a state s .”
Modelling Chemical Reasoning to Predict Reactions (2016, background): “29,38 Here, a binary reaction gets split into two half reactions, one for each reactant molecule, which also contain information about how the atoms and bonds of the reactant will be transformed and merged in the product.”