Large-scale analysis of metabolic networks: clustering metabolites by
their synthesizing capacities
University of Heidelberg
Humboldt University of Berlin
Last modified: June 23, 2006
We investigate the structure of metabolic networks by identifying sets of metabolites that have a similar synthesizing capacity. We measure the synthesizing capacity of a particular compound by determining all metabolites that can be produced from it, and call this set the scope of the compound.
To assess dissimilarities between scopes, we define a distance measure based on the Jaccard coefficient. We subsequently apply a hierarchical clustering method to identify groups of compounds with similar scopes. We find that compounds within the same cluster are generally chemically similar and often appear in the same metabolic pathway. For each cluster we define a consensus scope by determining the set of metabolites which is most similar to all scopes within the cluster. This allows for a generalization from scopes of single compounds to scopes of a chemical family (e.g. a certain group of amino acids). We find that only a few of the resulting consensus scopes are mutually disjoint while others overlap, and some consensus scopes are fully contained in others. Summarizing, our approach reveals a number of functional subunits of the metabolic network, and the size and overlap of their consensus scopes allow to arrange these subunits into a hierarchical setting.