Module sample correlation analysis in WGCNA

Question

[alim.hcu]

Dear All

I am doing WGCNA analysis of different tissue samples from a plant. After Module identification, I want to calculate module sample relationship analysis. For module and sample correlation, we have to use two different values one from module and another from samples. For module we can use expression of each module. But for samples, what type of data we will use for sample (leaf, root, stem, flower, sepal, petals).

Thank You

Answer 1

[Kevin Blighe]

The idea is to correlate or regress your phenotypes against the WGCNA modules. You can use anything that you want. If, for example, leaf is statistically significantly correlated to the pink module, then you would hypothesise that the genes in the pink module are related somehow to leaf

WGCNA is quite flexible and gives you, the user, much freedom to perform any analysis(es) that you please.

Kevin

[Shriyansh]

Hi Kevin Blighe thank you for your insights into such great details with WGCNA across the web.

I have a doubt regarding the type of analysis mentioned here.

Can we see how each module is associated with the treatment group as a measure of correlation? My question arises since the name of a 'treatment group' essentially is not a numeric value that can be used to calculate the correlation. (mentioned here: https://bioinformaticsworkbook.org/tutorials/wgcna.html#gsc.tab=0) from what I understand in the tutorial, the name 'corr' and values (-1,1) are the input given in the code, not the actual correlation function used. please let me know if I am understanding this correctly (or not).

also requesting to share some resources if there are any on the same. Thanks in advance!