Weird WGCNA plot

Question

[synat.keam]

Dear Seniors and All members,

I am wondering whether anyone has done weighted gene co-expression network analysis (WGCNA) from Bulk-RNAseq before. I followed the WGNCA tutorial using my data and the module detection outputs are here.

However, once I visualized the modules in clustering, I got weird output like the following

I have attached the clustering of modules from the tutorial, which looks nice and interpretable. I dont expect my data to be the same like this, but just would like to check whether I have done anything wrong with my code.

I have gone through identifying the soft-threshold power and following are the code chuncks I used. Wondering whether this is due to my data or might be sth wrong or missing there. Really appreciate any help from member who have done it before or anyone can help is welcome.

Thank in advance.

Kind Regards,

synat

[andres.firrincieli]

please, check the image

edit: also, define weird

[synat.keam]

Dear andres,

I got a bit of problem while I was writing my question and attaching images. I have now attached the full description of my inquiry. Hope you could have a look.

Thanks, synat

Answer 1

[andres.firrincieli]

Hi synat,

I have gone through identifying the soft-threshold power and following are the code chuncks I used. Wondering whether this is due to my data or might be sth wrong or missing there. Really appreciate any help from member who have done it before or anyone can help is welcome.

Is something that has to do with your data because a soft-thresholding power of 3 for unsigned network mean that you have a very strong driver of variation.

edit: If you run a PCA you will understand which samples are causing this problem

[synat.keam]

HI Andres,

You may be right. I found two outliers in my PCA plots. However so far I have not attempted to remove it yet because I do not have any proper reason to remove. However, I will try to remove the two outliers and run WGCNA again and will let you know how it looks. Thank for your prompt response. Much appreciated.

Thanks, synat