John A. P. Sekar

Department of Computational Biology
University of Pittsburgh School of Medicine

Poster: 
Title: Modular Construction Of Rule-Based Models

John A. P. Sekar and James R. Faeder

Abstract: Reaction models of biochemistry are characterized by dense and
interconnected networks of interactions. However redundancies exist in
the structure, kinetics and function of the participating molecule types
that can be exploited to provide a compact mathematical description of
the reaction network. Rule-based models exploit the redundancy that
arises from independent and modular structural elements within molecule
types. A second form of redundancy that can be exploited is that which
arises from very similar but subtly different molecule types that
warrants grouping of molecule types into type classes. Here, we extend
the rule-based framework to allow modules of reaction rules to be built
independently, reused and modified for type-redundancy and aggregated
automatically into a rule-based model with customized loading options.
This removes significant human effort from the construction process and
improves the scalability of construction and portability of rule-based
models.