[BioC] DESeq multi-factor export results

Ingrid Lindquist iel at ncgr.org
Wed Mar 21 16:08:11 CET 2012


Hi Simon,

See below.  If I'm not mistaken the padjGLM file has the adjusted pvals 
in the order that is in fit1, so I've been manually combining them.  Let 
me know though if this is not the correct assumption..
Ingrid

counts <- read.table ("pathtofile", header=TRUE, row.names=1)
icv_c_ctrl_Design<- data.frame(row.names=colnames(counts), condition= 
c("c", "c", "c", "c", "c", "ctrl", "ctrl", "ctrl", "ctrl"), libType= 
c("DSN-SS", "DSN-SS", "DSN", "DSN-SS", "DSN", "DSN", "DSN-SS", "DSN", 
"DSN-SS"))
library(DESeq)
cdsFull <- newCountDataSet (counts, icv_c_ctrl_Design)
cdsFull <-estimateSizeFactors(cdsFull)
cdsFull <-estimateDispersions(cdsFull)
fit1 <-fitNbinomGLMs(cdsFull, count ~ libType + condition)
fit0 <-fitNbinomGLMs(cdsFull, count ~ libType)
pvalsGLM <-nbinomGLMTest(fit1, fit0)
padjGLM <-p.adjust(pvalsGLM, method = "BH")
write.table(fit1, file="fit_c_ctrl")
write.table(padjGLM, file="padjGLM_c_ctrl")


On 3/21/2012 9:03 AM, Simon Anders wrote:
> Dear Ingrid
>
> On 03/21/2012 03:18 PM, Ingrid Lindquist wrote:
>> I am working with a multi-factor experimental design within DESeq and am
>> confused as to how to (when to) export my results. I am wondering at
>> what point can i write.table of my results to include both foldchange
>> and padj values for each gene.
>
> Please post your workflow with the commands you use so that I can 
> indicate where the results become available.
>
>   Simon
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at r-project.org
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: 
> http://news.gmane.org/gmane.science.biology.informatics.conductor


-- 

Ingrid Lindquist
National Center for Genome Resources
2935 Rodeo Park Drive East
Santa Fe, NM 87505
(505) 995-4426
iel at ncgr.org



More information about the Bioconductor mailing list