Composition of biochemical networks using domain knowledge

Tools

Goodfellow, Martin and Wilson, John and Hunt, Ela; (2010) Composition of biochemical networks using domain knowledge. In: COmputational Modeling in BIology NEtwork (COMBINE) 2010. Nature Precedings, GBR. (https://doi.org/10.1038/npre.2010.4966.1)

[thumbnail of npre20104966_1.pdf]
Preview
 PDF. Filename: npre20104966_1.pdf
Preprint
Download (5MB)| Preview

Abstract

Graph composition has applications in a variety of practical applications. In drug development, for instance, in order to understand possible drug interactions, one has to merge known networks and examine topological variants arising from such composition. Similarly, the design of sensor nets may use existing network infrastructures, and the superposition of one network on another can help with network design and optimisation. The problem of network composition has not received much attention in algorithm and database research. Here, we work with biological networks encoded in Systems Biology Markup Language (SBML), based on XML syntax. We focus on XML merging and examine the algorithmic and performance challenges we encountered in our work and the possible solutions to the graph merge problem. We show that our XML graph merge solution performs well in practice and improves on the existing toolsets. This leads us into future work directions and the plan of research which will aim to implement graph merging primitives using domain knowledge to perform composition and decomposition on specific graphs in the biological domain.

ORCID iDs

Goodfellow, Martin ORCID logoORCID: https://orcid.org/0000-0003-2151-8442, Wilson, John ORCID logoORCID: https://orcid.org/0000-0002-5297-657X and Hunt, Ela;
CORE (COnnecting REpositories)