Normalized transcript/exon counts for network analysis
1
0
Entering edit mode
maya.kappil ▴ 10
@mayakappil-18569
Last seen 21 months ago

Hi!

I would like to generate a cosplicing network similar to how it's described here (https://github.com/iancuo/cosplicingNetworks). My understanding is that I would need normalized exon/transcript counts prior to generating the network. For DEXSeq, I was planning to input the normalized exon counts from the DEXSEqDataSet generated as follows:

dxd <- DEXSeqDataSet(countData=count.data, sampleData=sample.data, design=~sample + exon + group:exon, featureID=counts(d)$featureid, groupID=counts(d)$geneid)

system.time({ dxd <- estimateSizeFactors(dxd) dxd <- estimateDispersions(dxd, quiet=TRUE) dxd <- nbinomLRT(dxd, reduced=~sample + exon) })

dxd_norm<-counts(dxd,normalized=T)

I was wondering whether a similar option is available to output normalized transcript levels from DRIMSeq?

Thanks!

Maya

drimseq dexseq wgcna sva • 421 views
1
Entering edit mode
@gosia-nowicka-9493
Last seen 6 months ago
Basel

Hello Maya,

no, there is no way to get normalized data from the DRIMSeq workflow. This is because DRIMSeq does not need to use normalized data when comparing proportions.

What you could do is to input your transcript counts into DEXSeq in the same way you did with exon counts.

0
Entering edit mode

Ah, ok - that makes sense! Thanks!!