Question: WGCNA module Eigengene PC1 and expression correlation
0
gravatar for Sri Dewi
15 days ago by
Sri Dewi0
University Medical Center of the Johannes Gutenberg University Mainz
Sri Dewi0 wrote:

Hi,

I am currently analyzing a small RNAseq data with 10 biological replicates for each 3 different timepoints (time1, time2, time3) from a mouse data. Firstly I mapped and quantified the data with miRDeep2 and proceed the counts data with DESeq2.

As suggested I used varianceStabilizingTransformation(DESeq.ds, blind=FALSE) to transform the counts as an input for WGCNA. Strangely if I plot the PC1 of the module eigengenes for each modules predicted, I got an opposite results regarding the expression of each timepoint.

In my raw expression data as well as in my DEG list, I could see that genes that are expressed lower in time1 have a higher PC1 (pink dots) and genes that are expressed higher in time3 have a lower PC1 (blue dots)

.

Can I interpret the PC1 value as an expression correlation of each time point? Or did I understood it all wrongly? I have rather only weak knowledge about statistics in general.

Thanks in advance for any kind of help and hints!

Dewi

deseq2 wgcna rna-seq • 91 views
ADD COMMENTlink modified 15 days ago • written 15 days ago by Sri Dewi0
Answer: WGCNA module Eigengene PC1 and expression correlation
0
gravatar for mikhael.manurung
15 days ago by
Netherlands
mikhael.manurung200 wrote:

I could not fully understand what you mean by strange or opposite results (your image link is broken). However, module eigengene (ME; PC1) indeed summarises a module's activity. Comparing raw expression values and ME is like apples and oranges, by the way. I would'nt be surprised that your ME showed a 'different' result simply because your gene of interest is somehow affected by other genes in the respective module.

ADD COMMENTlink written 15 days ago by mikhael.manurung200

Sorry for the broken link, hopefully it will work this time. https://cdn1.imggmi.com/uploads/2019/11/28/9fa736504f4c1eb2207db40a262f4091-full.png

And also sorry for the confusion, what I meant was the normalized counts (RPKM). But I did found out the problem with my code, instead of using signed network as I thought I did in my script, I used unsigned networks that brought out the unexpected results.

Thanks for your help!

ADD REPLYlink written 14 days ago by Sri Dewi0
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 149 users visited in the last hour