Question: Multiple transcription factors over expression analysis
gravatar for sebastiano.curreli
3.5 years ago by
sebastiano.curreli0 wrote:

Dear community,

I'd need some help in the analysis of my RNAseq data using DEseq2.

I'm trying to characterize the effect of the over expression of three transcription factors (x,y,z) on the transcriptome a cell line.

I have the following groups and each contains 5 replicates.

1) Ctrl

2) x

3) y

4) z

5) x,y,z

As far as I understood I can set up my analysis as a "single factor" with "multiple levels" (as in explained in the manual 3.2 paragraph).

So, if I understood correctly, in order to obtain the transcriptional changes induced by each transcription factor I would need to run the following code:

condition<-c("x","x","x","x","x","y","y","y","y","y","z","z","z","z","z","xyz","xyz","xyz","xyz","xyz","ctrl","ctrl","ctrl","ctrl","ctrl") #declare my conditions
ddsHTSeq <- DESeqDataSetFromHTSeqCount(sampleTable = sampleTable, directory = directory,design= ~ condition) #building the dds object
ddsHTSeqfilt <- ddsHTSeq[ rowSums(counts(ddsHTSeq)) > 1, ] #filtering 0 values
ddsHTSeqfilt$condition <- relevel(ddsHTSeqfilt$condition, ref="ctrl") # specifying to use ctrl as reference
dds <- DESeq(ddsHTSeqfilt)
DEx<-results(dds, contrast=c("condition","ctrl","x")) #extracting the differentially expressed genes between ctrl and "x"

Can you please confirm that what I'm obtaining in this way are the differentially expressed genes between "x" and "ctrl" groups?

Thanks in advance for your help.




rnaseq deseq2 • 683 views
ADD COMMENTlink modified 3.5 years ago • written 3.5 years ago by sebastiano.curreli0

Seems about right to me, but you probably want to swap the order of the levels in your contrast argument, eg: contrast=c("condition", "x", "ctrl")) so you get ctrl in the denominator of the logFoldChange you calculate.

ADD REPLYlink modified 3.5 years ago • written 3.5 years ago by Steve Lianoglou12k

Thanks for your prompt suggestion Steve.

ADD REPLYlink written 3.5 years ago by sebastiano.curreli0
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 205 users visited in the last hour