Constructing Enzyme Network

Question

[Woa]

I wish to construct an Enzyme network, where two enzymes would share an edge, if the product(small molecule) of one enzyme is used as a reactant for the other enzyme.

I've a couple of hundreds of such enzymes for an organism, and I know their EC numbers. Can somebody provide a brief guidance about where to start from? Can the API of KEGG be used? If so, can one cite a small example?

Thanks in advance

WoA

Answer 1

[Alex Paciorkowski]

You can use Cytoscape to do this. If you have a tab-delimited text file with the following information:

EC 1.2.4.1     shares_reactant_with       EC 1.1.1.27

You can then import the file into Cytoscape, select the select the columns "source interaction" for column 1, "interaction type" for column 2, and "target interaction" for column 3. This will draw a network, which you can modify using Cytoscape's layout options. Is this what you are after?

[Woa]

Thanks Alex, but I need to establish the edge relationship first. I just have a collection of nodes(enzymes) with the EC numbers. I need to know which reaction they catalyze, and based on that info. I have to tell whether they can have an edge between them. Their can be some potential glitches. Common co-factors like FAD, ATP,ADP etc. I think should be excluded and enzymes separated in subcellular compartments where they have least chance to react should be excluded. Anyway that sort of refinement comes second to finding out the catalyzed reaction, the reactants and products.

[Alex Paciorkowski]

Then I would recommend using a structured ontology like KEGG's BRITE hierarchy, that will allow you to identify (and exclude) the common co-factors you mention. http://www.genome.jp/kegg/brite.html

I'm not sure however whether all of your enzymes are represented in the hierarchy.