Entering edit mode
20 months ago
Penny
•
0
Hi Im a newbie in Bioinfo field. I am now doing WGCNA and DEG on RNA-seq data.
I was wondering if DE genes should be the same as hub genes identified through WGCNA? Actually I am a bit confused about the different purposes between WGCNA vs DEG.
In my analysis, the hub genes in the significant WGCNA module (I have only one significant module) are not the same as the DEG I got from DE analysis. My hub genes were determined based on module membership (p<0.05 and the top 10 highest correlation coefficient).
Many Thanks if any one could enlighten me on this.
Hub genes are highly connected genes based on correlation metrics such as Pearson. DEGs are genes with statistically significant differences in counts between two or more groups. That's simply two completely different concepts.
Hi ATpoint, thanks a lot for your instant response! This is my first time using the website.
Just to make sure I understand right. So what WGCNA analysis does is to spread out all the gene counts (in regardless of the groups) and check which module is highly related to the trait(s)? It is a correlation between genes (modules) and traits.
And DE is to check the diff between groups (eg., WT vs mutant mouse)?
Thanks in advance.
I'm doing WGCNA for the first time and found the following paper very helpful: https://bmcbioinformatics.biomedcentral.com/articles/10.1186/1471-2105-9-559
WGCNA aims to build a network (aka graph) using an adjacency matrix based on measures of correlation in the expression of genes in your sample set. One of the questions you can ask with WGCNA is whether any, and if so which, clusters are associated with sample traits or experimental design.
From the above linked paper:
Hi jv, thank you so much for the detailed info. I just downloaded the paper you recommended. The last sentence of your quote from the paper kinda gives me an aha moment. It enlightens me on the general main purpose of WGCNA. I will read through the paper. Thanks again!