Section: Application Domains

Abstraction of Biological Cell Signaling Networks

Keywords: Biology, Health, Static analysis.

Protein-protein interactions consist in complexations and post translational modifications such as phosphorilation. These interactions enable biological organisms to receive, propagate, and integrate signals that are expressed as proteins concentrations in order to make decisions (on the choice between cell division and cell death for instance). Models of such interaction networks suffer from a combinatorial blow up in the number of species (number of non-isomorphic ways in which some proteins can be connected to each others). This large number of species makes the design and the analysis of these models a highly difficult task. Moreover the properties of interest are usually quantitative observations on stochastic or differential trajectories, which are difficult to compute or abstract.

Contextual graph-rewriting systems allow a concise description of these networks, which leads to a scalable method for modeling them. Then abstract interpretation allows the abstraction of these systems properties. First qualitative abstractions (such as over approximation of complexes that can be built) provide both debugging information in the design phases (of models) and static information that are necessary in order to make other computations (such as stochastic simulations) scale up. Then qualitative invariants also drive efficient quantitative abstractions (such as the reduction of ordinary differential semantics).

The work of the Abstraction project-team on biological cell signaling networks ranges from qualitative abstractions to quantitative abstractions.