Coexpression analysis

Question

[priya.bmg]

Hello

I am using R package, CEMi tool for coexpression analysis. My expression data set has 34 rows and 525 columns. When I run the command for full gene expression analysis, I get an error with message less than 3 rows

cem <- cemitool(expr_ch)

Error in WGCNA::pickSoftThreshold(expr_t, powerVector = powers, networkType = network_type,  : 
  The input data data contain fewer than 3 rows (nodes).
This would result in a trivial correlation network`enter code here

Has anyone encountered this issue

Thanks

[andres.firrincieli]

My expression data set has 34 rows and 525 columns.

cemitool require a data.frame containing expression data genes in the rows and sample names in the columns.

What is the output of dim(expr_ch)?

[priya.bmg]

34 524

[andres.firrincieli]

So you have 34 genes and 524 samples?

cemitool require a data.frame containing expression data genes in the rows and sample names in the columns.

[priya.bmg]

Yes, have 34 genes and 524 brain regions (samples)

[andres.firrincieli]

have 34 genes and 524 brain regions (samples)

Are you working with an expression matrix of differentially expressed genes?

Here is what CEMiTool authors say in the manuscript:

Prior to identifying co-expression modules, it is recommended to filter input genes by either mean expression or variance, rather than by differential expression since this would invalidate the scale-free topology assumption.

If you are using the DEG this could explain the error:

Error in WGCNA::pickSoftThreshold(expr_t, powerVector = powers, networkType = network_type,  : 
  The input data data contain fewer than 3 rows (nodes).
This would result in a trivial correlation network`enter code here