Biological Sciences, Chemistry
Cell and Molecular Biology
Metabolic networks; network theory; power law distribution
Inspired by the landmark paper “Emergence of Scaling in Random Networks” by Barabási and Albert, the field of network science has focused heavily on the power law distribution in recent years. This distribution has been used to model everything from the popularity of sites on the World Wide Web to the number of citations received on a scientific paper. The feature of this distribution is highlighted by the fact that many nodes (websites or papers) have few connections (internet links or citations) while few “hubs” are connected to many nodes. These properties lead to two very important observed effects: the so-called small world property and robustness to random attacks. The small world property holds that although the size of e.g. the World Wide Web is massive, the distance between any two sites is small (19 clicks separate any two websites). The robustness feature is a result of the networks having proportionally few hubs; a random attack (taking down a random website) is not likely to hit a hub like Facebook or Twitter.
In biology, metabolic networks are constructed by linking metabolite nodes to one another if there is a biochemical reaction that converts one to another. The degrees of each node in these networks have been suggested to follow a power law distribution. In this project, we will use the PSAMM software to analyze metabolic networks in a collection of curated metabolic models from the literature with an aim to determine whether the degrees of nodes in each metabolic network follow a power law distribution. We will investigate how the different representations of the networks may influence the power law fitting. We will also compare the power law fittings between models across different species and note how similar or different the structure of the species’ networks are based on this specific feature.