Integrate several layers of biological information to build a biological network

Question

[lessismore]

Dear all,

i am searching for uptodate programs/packages to integrate several data sources (e.g. RNA-Seq, metabolic, protein-protein interaction, etc.) in the same network in order to give a weight to the undirected edges based on the provenience of the gene interaction information.

On the other hand if you know programs/packages to first build the co-expression network and then integrate the other data sources in order to weight the interactions in a second phase would be also useful.

Another thing i would find useful to transform a undirected network in a directed network based on the same principle i wrote above.

Any tips, articles, packages, platforms will be highly appreciated.

Thanks in advance

[Joe]

First thing that comes to mind is STRING, though it may not be exactly what you're looking for.

It identifies networks of co-occuring genes etc (for example in differential expression experiments), but the strength of the edges is determined by other factors such as genes that are mentioned together in journal articles and so on as well. It does integrate many layers of information though.

[Kevin Blighe]

This is an active area of research and there are, as yet, no concrete solutions. It was a hot topic in the lab where I was recently based as postdoc in Boston ,USA. So, feel free to explore solutions.

Answer 1

[bharata1803]

Several network source that I know:

WGCNA : build a coexpression network

STRING database : provide gene and protein interaction based on mention in the same published paper

KEGG pathway : a pathway network from manually curated data

http://regulatorycircuits.org/ : builds the network based on pairing transcription binding matrix in the form of PWM into enhancer and promoter region annotation data available from FANTOM5 database

http://www.regnetworkweb.org/ : interaction among regulatory molecules, both during transcription process and post-transcription process