voom on normalised counts
1
0
Entering edit mode
i.sudbery ▴ 40
@isudbery-8266
Last seen 4 weeks ago
European Union

Hi,

We wish to do exon analysis on an existing RNA-seq dataset - we actually only need to do it for one gene. However, the only publicly accessible data is pre-normalised using a quantile normalisation. I know that this data would be unsuitable for DEXSeq (or DESeq/edgeR), but would it be valid to use limma voom on the data? I don't know what statistical assumptions voom makes.

Cheers,

Ian Sudbery

 

limma voom rnaseq • 4.1k views
ADD COMMENT
0
Entering edit mode

Quantile normalized counts? That sounds very strange. Or are they actually RPKM values? Please describe the pre-normalised expression quantities you have in more detail.

ADD REPLY
2
Entering edit mode
Aaron Lun ★ 28k
@alun
Last seen 7 hours ago
The city by the bay

voom requires counts as input, in order to compute sensible log-CPM values and to fit a sensible mean-variance trend. So, if you don't have counts, then the best you can do would be to log-transform the data (if that hasn't already been done) and analyze those values as if they were microarray intensities with limma. See A: Differential expression of RNA-seq data using limma and voom() for details.

P.S. limma and voom should be separate tags, otherwise watchers of either tag won't get notified.

ADD COMMENT
0
Entering edit mode

And use trend=TRUE when running eBayes() in limma. In other words, use limma-trend.

ADD REPLY

Login before adding your answer.

Traffic: 767 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6