Rna Raw Count Data Normalization

Question

[jack]

Hi,

I have raw count data of RNA seq. I want to normalize it. Which normalization methods do you recommend and which R packages I can use?

Answer 1

[Devon Ryan]

DESeq2, edgeR and limma (with voom()) are the common choices. They all have default normalization methods that you might as well stick with unless you know enough to judge how well they're working (which, if you're asking this question, is presumably not the case). I personally prefer DESeq2, but some designs are more easily accomplished with limma.

[jack]

used DESeq2, and I want to do log transformation. I've got this error. what should I do? rld <- rlogTransformation(my_rawcount, blind=TRUE)

the error is : Error in (function (classes, fdef, mtable) : unable to find an inherited method for function ‘sizeFactors’ for signature ‘"matrix"’

[Devon Ryan]

You're giving a matrix to a function that expects a DESeqDataSet object. Try creating the appropriate object first, likely with the DESeqDataSetFromMatrix() function.