This is a test version of Biostars. For the public version, visit https://www.biostars.org.
WGCNA for less than 15 samples

In FAQ's of WGCNA package it's mentioned "We do not recommend attempting WGCNA on a data set consisting of fewer than 15 samples". What could be an alternate approach if one have let's say 6 samples? Can anybody explain if for example I want to analyze this gene expression data then from where will I start?

wgcna
• 4,539 views • link
updated by Ram Ram • written by Bioinformatist Newbie Bioinformatist Newbie

What is your actual goal?

• 0 views • link
written by Devon Ryan Devon Ryan

1 answer

You can make an easy co-expression network by simply constructing the symmetric correlation matrix of all genes (adjacency matrix as its called in WGCNA) then filtering that matrix by a hard threshold such as r = 0.70.

Although, based on your newest comment I am not sure why differential expression analysis is not sufficient. Couldn't they be ranked on highest absolute fold change or lowest FDR adjusted p-value?

• 0 views • link
updated by Ram Ram • written by Brian Gudenas Brian Gudenas

Aren't those attempting to answer two different questions?

With differential analysis, we're looking at how gene expression changes between conditions.

In making a correlation matrix, is this not a means of realizing relationships between genes, regardless of fold change across experimental conditions?

I'm asking as a novice.

• 0 views • link
updated by Ram Ram • written by rebeliscu • rebeliscu

Log in to answer this question.

More posts like this