DESeq: defining conditions

[mouchkam]

Hello,

I'm very new to DESeq and am hoping to verify that my script is correct, specifically relating to importing my data (a text file with raw count data for two treatments, AC and SC, 4 biological replicates for each) and defining my treaments. Below is my script and I've attached a txt file with the first few lines of my data as an example of my input format. Specifically, I'm wanting to compare AC to SC. The script seems to work beautifully, but I just want to check that I haven't incorrectly imported the data and/or assigned the treatments.

Thanks so much!

CountData<- read.table("AC_SC_readCountsDESeq.txt",row.names=1, header=TRUE,)
conds <- c(rep("AC",4), rep("SC",4))
cds <- newCountDataSet( CountData, conds )
LibrarySize <- estimateSizeFactors( cds )
sizeFactors( LibrarySize )
head( counts( LibrarySize, normalized=TRUE ) )
Disp <- estimateDispersions( LibrarySize )
DE <- nbinomTest( Disp, "AC", "SC" )

[sdriscoll]

looks good to me. i usually don't make all of the extra variables but it doesn't matter. what i mean is if i had run it i would have kept redefining 'cds' since each of the commands only updates the count dataset object with new/updated information and nothing is lost.

so it would look like this...

CountData<- read.table("AC_SC_readCountsDESeq.txt",row.names=1, header=TRUE,)
conds <- c(rep("AC",4), rep("SC",4))
cds <- newCountDataSet( CountData, conds )
cds <- estimateSizeFactors( cds )
sizeFactors( cds )
head( counts( cds, normalized=TRUE ) )
cds <- estimateDispersions( cds )
DE <- nbinomTest( cds, "AC", "SC" )

[mouchkam]

Wonderful! Thanks for your help and for your suggestion. I originally thought that reusing the same variable was confusing, but since it seems like pretty standard practice, I think I just confuse myself when altering code if I define new variables. Thanks again!