Using Rules & Pathway Databases to Create Quantitative Mechanistic Models in Virtual Cell

Abstract

Recent technological advances have resulted in an accumulation of experimental data and a growing interest in using this data to build quantitative models of biological processes. The traditional approach to build such models involves manual formulation of model hypothesis and data search. This is becoming inadequate due to the degree of complexity in the models and the great variety of data available. In particular, for understanding and modeling the dynamics of protein-protein interactions, the mechanisms of interactions have to be described at the level of protein sites, the parts of proteins that are responsible for protein-protein interactions, such as protein phosphorylation sites and interaction domains. The rule-based approach provides modelers with an opportunity to efficiently use such information. However, despite the high relevance of the site-specific details of protein-protein interactions for understanding system behavior, rule-based models incorporating these details are not very common, because of difficulties in mining and using this information for modeling. To address this need, we developed 2 new capabilities within the Virtual Cell modeling and simulation framework. The first one is BioNetGen@VCell, which enables the user to create rule-based models and combine rules and reactions in a single interface. The other, PathwayCommons@VCell, enables the user to easily extract information from external pathway databases and create computational models of pathways. We describe the technology underlying these tools and present an example model that makes use of them.