COMPUTER SCIENCE COLLOQUIUM

PhD colloqium: Analysis of Generative Chemistries -- Graph Grammars, Hyperflows and Composition of Rules

Jakob Lykke Andersen Department of Mathematics and Computer Science University of Southern Denmark, Denmark

Tuesday, 27 November, 2012 at 14:15 IMADA's Seminar Room

ABSTRACT

We use labelled undirected graphs to model molecules and transformation rules in the Double Pushout formalism to model chemical reactions. A set of starting graphs and transformation rules, a graph grammar, implicitly defines a directed hypergraph with derivable graphs as vertices and the direct derivations as directed hyperedges. On these hypergraphs (reaction networks) we define a hyperflow model in order to find interesting chemical pathways and specific pathway motifs, e.g., autocatalysis.

For exploring graph grammars directly we define a type of composition of transformation rules. This is used to infer higher level rules which represent chemical properties. We also present a general strategy framework for (chemical) graph transformation based on rule composition.

Host:

---

SDU HOME | IMADA HOME | Previous Page

---

Daniel Merkle