Scale Free Topology Fit In Wgcna

1

Entering edit mode

11.2 years ago

pixie@bioinfo ★ 1.5k

I have analyzed this dataset (GSE26280) using NCBI GEOtoR . I wanted to perform WGCNA analysis on the differentially expressed genes. But the scale free topology fit is coming very different ( and also it starts from a -ve value). Is it because of the sample size ? I have taken 12 samples in this case. Any explanation for this graph would be really helpful.

enter image description here

network r • 5.4k views

ADD COMMENT • link 11.2 years ago by pixie@bioinfo ★ 1.5k

1

Entering edit mode

I remember seeing similar "off" results when I accidentaly worked on the transposed expression matrix. If I remember correctly, WGNCA expect a samples * genes matrix, is this how your expression data is laid out?

ADD REPLY • link 11.1 years ago by Hayssam ▴ 280

0

Entering edit mode

No, my gene expression has always been genes*samples. I just pass the transpose around everywhere.

ADD REPLY • link 11.1 years ago by s.ritchie73 • 0

0

Entering edit mode

That's bizarre. Can you plot an image of the density of the similarity matrix?

ADD REPLY • link 11.2 years ago by s.ritchie73 • 0

Login before adding your answer.