Question

Analysis small RNA data with DESeq2

0

Entering edit mode

zcc_1127 • 0

@zcc_1127-12969

Last seen 7.0 years ago

Dear all,

I have some problems about small RNA sequencing data analysis.

After mapping to the genome, I calculate the raw counts in the regions I interested in, so I get a matrix like this:

		untreated1	untreated2	untreated3	treated1	treated2	treated3
region1	Chr1:200-400	x	x	x	x	x	x
region2	Chr2:600-800	x	x	x	x	x	x
region3	Chr2:300-900	x	x	x	x	x	x

My questions are:

1. Can I use DESeq2 to do differential expression analysis of these small RNA clusters? My regions are not cover the whole genome.

2. Is it necessary to input all the raw data when I do normal differential analysis, or I can just input a subset of the raw data?

3. Can I setup the SizeFactor of each library befor nomalizition?

Thanks for your help!

Chengcheng

deseq2 differential gene expression • 1.3k views

ADD COMMENT • link updated 7.0 years ago by Michael Love 41k • written 7.0 years ago by zcc_1127 • 0

score 0 · Answer 1 · 2017-05-05

hi,

To the extent that your data is similar to the typical RNA-seq dataset, you can use DESeq2, I can't say much more than that from what you've provided. I think the important things to consider are that the DESeq2 model requires that some of the features are not differentially expressed. So you shouldn't subset to the features that you think are differentially expressed, or that DE signal will be removed during normalization. Size factor estimation is the normalization step that takes place by default. If you provide size factors, via

sizeFactors(dds) <- x

...then those will be used for normalization and DESeq() will not attempt to estimate size factors.