Entering edit mode

8 weeks ago

Ankit
▴
320

Hi everyone,

I have a matrix as in example and I want to calculate the variability across the rows.

Please suggest me which test to use and which function in r can do this?

If there is code anyone can share will be much more helpful.

I appreciate the help.

Thank you Ankit

What kind of data are this?

Methylation arrays-betavalues

cgs are basically single cpg sites

Limma: https://bioconductor.org/packages/devel/workflows/vignettes/methylationArrayAnalysis/inst/doc/methylationArrayAnalysis.html

Thanks

A function in Limma link (varFit) is calculating variability between two groups. But I am looking for the most variable CpG sites across samples. Simply, if the methylation value of CpG (lets say cg1) vary among sample1 to 7 then it is the variable CpG and I need statistical way to do include/exclude it. I do not have any group. These all samples are controls.

Any other approach?

I noticed the post about rowVar. The issue is how to select the threshold of variance to classify my most variable regions.