Quantile normalization and DESeq

[dideco]

Dear all,

I'm looking to detect differentially expressed genes between different conditions, each represented by minimum 3 replicates. Preliminary in-depth analysis of my data showed inconsistent distributions of read counts between my samples. Normally I would do DE analysis using the EdgeR package and thus the TMM normalization method. However, EdgeR assumes comparable distribution of read counts, which I have not.

Looking to other ways to normalize my data I regularly came across the workflow to quantile normalize with the LIMMA package function, and then using the DESeq package for DE analysis, manually specifying the sizeFactors to accomodate for the quantile normalization. However, it is unclear to me how this is exactly done? Is this uberhaupt a valid way to do this?

Any help would be much appreciated!

Thanks,
Dieter

[archana2287]

i want to perform the quantile normalization . Is it possible to do this Deseq , EdgeR ??

[dpryan]

Neither edgeR nor DESeq support quantile normalization. Anyone who said that they did that didn't know what they were talking about. If you want/need to quantile normalize, use limma/voom.

[archana2287]

Thank you for your suggestion.