Question: After WGCNA analysis: Gene regulatory Network ?
gravatar for Gabin
4.7 years ago by
Gabin20 wrote:

Dear all,

I am analyzing a microarray dataset composed by 32 samples and 50.000 genes. My goal is to create some gene networks, more precisely, I want to know the interaction type between genes (activation/inhibition for example)

I found some interesting modules using the WGCNA package. In each module, genes are more or less highly connected and I want to discover the nature of this connection.

Working on each module separately to build one gene regulatory network by module was my idea.

The question is : Is this method biologically or mathematically relevant ?

Thanking you in advance.

Best regards

Gabin Pierlot

network wgcna • 2.9k views
ADD COMMENTlink modified 4.7 years ago by ivivek_ngs5.0k • written 4.7 years ago by Gabin20
gravatar for WouterDeCoster
4.7 years ago by
WouterDeCoster44k wrote:

You could check for biological relevance by performing enrichment analysis (e.g. Enrichr, webgestalt, gorilla, panther) on the obtained clusters.

ADD COMMENTlink written 4.7 years ago by WouterDeCoster44k
gravatar for ofisher
4.7 years ago by
ofisher0 wrote:

To identify modules that may be of more relevance, after identifying them (currently under review) we recently performed enrichment analysis to see if the modules made biologic sense. In our case they did and we then went on to characterizing the networks with highest biologic relevance to our study.

ADD COMMENTlink written 4.7 years ago by ofisher0
gravatar for ivivek_ngs
4.7 years ago by
Seattle,WA, USA
ivivek_ngs5.0k wrote:

There are different ways to do a network analysis. Most viable way is to do GSEA as others are suggesting to find a meaning biological modules from the given modules of WCGNA. What you can do is take each module make the GO enrichment analysis or any standard pathway analysis and then try to score the pathways based on p-value or FDR values and try to find the top pathways that are of convergent biological sense to your phenotype or experimental design. If it does then you can use any Cytoscape to create the network visualization for that module that is at the highest priority with significant p or FDR value giving you better biological relevance.

ADD COMMENTlink written 4.7 years ago by ivivek_ngs5.0k
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1509 users visited in the last hour