Converting KEGG metabolic pathway network to gene network

Question

[ptdikshi]

Hi,

I wish to study the impact of SNP data on the metabolic pathway network, using graph theory. For this, I want to map the KEGG metabolic network to a gene network (enzymes -> proteins -> genes) . Has anyone worked on this before? . Would appreciate any pointers! . TIA!

[ivivek_ngs]

Are you particularly interested in only KEGG? There are tools like Reactome, IPATH2 or even IPA or Pathwaycommons where you can feed a list of genes and restrict to your desired pathway and then create a network. However if you have specific metabolic pathways from your SNPs you can extract the gene list and then try to use string to get the interactions and finally use informations from these interactions and plot a network using Cytoscape. Are you trying device an algorithm of your own to find networks of metabolic pathways from SNP data? Then it is different. I am just trying to say about common usage stuffs. Alternatively you can also take a look at this paper if its helps.

[ptdikshi]

You are right. I am trying to devise an algorithm to find 'sub-graphs' within the metabolic pathway network, using the SNP data. Since KEGG already provides the pathway to gene mapping for all metabolic pathways, I thought it would be possible to get the enzyme to gene mapping from KEGG, so as to be able to represent the metabolic pathway network as a 'gene network'. Am open to using any other tools.

[ivivek_ngs]

It will be a whole new thing on its own then, am not familiar with devising sub-graphs in metabolic network through programmatic way. I usually create my network using the tools I have mentioned but never used a programmatic approach. So may be some one else can say. How about the paper I linked? Does it ring in some ideas? But some digging gave me the below hits, which might be useful for you

1. iSubpathwayMiner in R or another link for the same
2. How To Map Kegg Data - Pathways As Nodes, Compounds As Edges.

Good luck!

Answer 1

[Girolamo]

You can convert the human enzyme to kegg entries using: http://rest.kegg.jp/link/hsa/enzyme then you can retrieve the pathway in which you are interested in with: http://rest.kegg.jp/link/hsa/pathway and then map the pathway id to the description: http://rest.kegg.jp/list/pathway/hsa

[ptdikshi]

This would work well!. The enzyme link maps enzyme entries to gene entries. The pathway link maps pathway names to gene entries. Thank you!