Question: Why does my TOM network heatmap for WGCNA display red background and golden modules, showing opposite colors to the Tutorial?
0
gravatar for kingwind333
11 months ago by
kingwind3330 wrote:

Dear guys, I tried to replicate the WGCNA tutorial step by step (https://horvath.genetics.ucla.edu/html/CoexpressionNetwork/Rpackages/WGCNA/Tutorials/FemaleLiver-05-Visualization.pdf) Therefore I used the code exactly the same as the tutorial. However I got the TOM heatmap like this, with red background and golden modules. It's so wired that the colors are opposite to the figure in the tutorial. I googled a lot and found no similar phenomenon. How could I get this TOM heatmap like the tutorial? Thanks very much. code and plotenter image description here

![the TOM plot with red background and golden modules][3]

ADD COMMENTlink modified 8 months ago by derekterryandhusky0 • written 11 months ago by kingwind3330
remove.packages('WGCNA')

restart R

BiocManager::install('WGCNA')

reload data and redo TOMplot() still the red figure

ADD REPLYlink modified 11 months ago • written 11 months ago by kingwind3330

In TOMplot() you can use additional arguments such as those of the heatmap() function. Use the 'col' argument to set the color palette of the heatmap

ADD REPLYlink modified 11 months ago • written 11 months ago by andres.firrincieli620

Thanks. I still don't know why I need to use 'col' argument while the others including the tutorial does not. Could you please give detailed code example that how to use 'col' argument in TOMplot()?

ADD REPLYlink written 11 months ago by kingwind3330

I have been encountered the same problem. But I guess the aforementioned solution, which just simplily revert the color mapping, may not be correct.

I did inspect the TOM matrix and TOMplot function. The former consists of a larger proportion for Bigger Number(such as 0.8+ 0.9+),and The latter.,I found , call the built-in function — heatmap to draw the results, which will map the higher value with dark color and lower one with light color.

Any advices welcomed ! I have been stucked for moren than one month.

ADD REPLYlink written 8 months ago by derekterryandhusky0

Hi Derek,

In plotTOM() you are plotting a dissimilarity matrix dissTOM = 1-TOM

ADD REPLYlink written 8 months ago by andres.firrincieli620
6
gravatar for andres.firrincieli
11 months ago by
andres.firrincieli620 wrote:

Try this

library(gplots)
myheatcol = colorpanel(250,'red',"orange",'lemonchiffon')
TOMplot(plotTOM, geneTree, moduleColors, main = "Network heatmap plot, all genes", col=myheatcol)
ADD COMMENTlink written 11 months ago by andres.firrincieli620

It worked! But before your useful advice, I never changed 'colorpanel' to something like

'myheatcol = colorpanel(250,'gold',"orange",'darkred')'

I mean, if now I set the colorpanel like (250, 'gold',"orange",'darkred'), I can get that blood-smear-like figure again, not exactly the same, which is that the golden module is not as bright as the figure with default 'col'. Does this mean that the figure got by set the col = colorpanel(250,'red',"orange",'lemonchiffon'), also shows some different details from the tutorial figure? enter image description hereenter image description here Anyway, problem solved,,, almost. I will go on trying to find what's wrong with the default argument in my TOMplot(). Thank you so much!

ADD REPLYlink written 11 months ago by kingwind3330
1

I could be wrong but If you followed every step of the tutorial, I would not bother too much about the default color palette used in TOMplot(). It could be a version problem. Those tutorial were made in 2014

ADD REPLYlink written 11 months ago by andres.firrincieli620

Yes u r right. Then I won't bother either. Thank you!

ADD REPLYlink written 11 months ago by kingwind3330

for my opion, this step just reverts mapping process, in wich higher value tends to be "lemonchiffon" (ligher), while lower value tends to be "darked". So it's still hard to address the paradox: different modules have higher similarity than memebers in one module.

ADD REPLYlink written 8 months ago by derekterryandhusky0

There is no paradox. In TOMplot() you are plotting a distance matrix distTOM = 1-TOM. Therefore, genes in different modules have higher dissimilarity

ADD REPLYlink written 8 months ago by andres.firrincieli620
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 2.3.0
Traffic: 1525 users visited in the last hour