WGCNA - how to interpret gene cluster dendogram

Question

[janinu]

Hi,

I am a bioinformatics newbie trying to figure things out. I need help understanding what information I am supposed to gleam from a WGCNA cluster dendogram, such as the one below. I more or less understand how to read it I think, but I don't understand what useful information there is here. I can see that the dark blue and the green (merged) modules are those with the largest number of genes, but what else am I supposed to understand?

Also, I don't understand how the color panels at the bottom don't have clean blocks of color. For example, in the dark blue block of color on the bottom panel, we have some streaks of grey, indicating genes that belong to the grey module. How can they be situated in between genes that belong to the dark blue module?

Any help is much appreciated!

Answer 1

[andres.firrincieli]

The overlapping colors in the gene dendrogram is explained in this post: link. The supplementary material can be found at this link

I can see that the dark blue and the green (merged) modules are those with the largest number of genes, but what else am I supposed to understand?

On the left portion of the dendrogram most genes do not cluster very well. There is a lot of noise (grey color), and noisy genes usually populate the upper part of the dendrogram. If you lower the cutHeight parameters or increase the minimum module size (minClusterSize) in cutreeDynamic, you will see that most of the genes in the left portion of the dendrogram will be part of the grey module.

edit: More info about Dynamic Tree Cut algorithm can be found here: link