WGCNA Timepoint Analysis

Question

[singlecell_bio]

Hello! I am working on a WGCNA analysis to find co-expressed genes between different cell types. I created multi expression matrix based on timepoints to find co-expressed genes. However, when I use the plotDendroAndColors function, I get an error that says " object 'moduleColor' not found."

I also wanted to know the minimum amount of samples required for the WGCNA analysis. Because, when I try to compare data of two replicates of a cell type with two other replicates of a different cell type, the function returns an error saying that the number of genes isn't enough.

Any other advice on how to go about the WGCNA consensus analysis will also be much appreciated.

Thank you!

Answer 1

[Kevin Blighe]

Hi,

The first error means that you have no object in your workspace named moduleColor. Please take a look again at the tutorial(s): https://horvath.genetics.ucla.edu/html/CoexpressionNetwork/Rpackages/WGCNA/Tutorials/

As you have not shown any reproducible code, we cannot debug this further for you.

With respect to the minimum number of samples, that is difficult to answer; however, it forms part of the FAQ written by the developer: How many samples do I need?

Kind regards, Kevin

[singlecell_bio]

Thanks, Kevin! I have one more question. What is the difference between the network analysis and the consensus analysis? Does it mean that analysis within a group is network analysis and between-group analysis in a consensus analysis?

[Kevin Blighe]

The consensus module detection, to WGCNA, is about processing 2 separate datasets together, I think, and then finding consensus modules between each. I guess that we could technically analyse a Disease and Control group via this approach. I have not tested it extensively.

Normally, I would build separate graph objects for each group, and then join these separate networks via hub genes, or just analyse them separately and look at things like vertex degree, hub score, etc.

There are no real standards in network analysis.