Applying WGCNA in multiple data types

Question

[rin]

Hi all

As the newbie I am, I would really appreciate your help!

I would like to apply a similar approach to WGCNA, but taking into account more than one omics data types, so I cannot use WGCNA itself (I think).

I was considering doing the following:

1. Instead of the co-expression measure, I will create a matrix where I will add up the correlation values that resulted in three independent correlation tests in each of the data type.
2. Convert the above matrix into an adjacency matrix, if the summed correlations are over a threshold (I was considering a cut off of 0.7 considering that for some of the data types I am using most of the correlations are low).
3. Use WGCNA´s functions for the rest of the analysis.

Would that be a reasonable thing to do with respect to the first steps? I have not found any similar approaches, or cases where they have used WGCNA to something other than co-expression.

Thanks you in advance.

R.

Answer 1

[Kevin Blighe]

You could follow the methodology of a recent work in which I was involved, whereby WGCNA was used to integrate gene expression and metabolomics data. Please see: An Integrative Transcriptomic and Metabolomic Study of Lung Function in Children With Asthma. .

Kevin