DESeq2 - batch, condition and tissue information
2
0
Entering edit mode
Ela__77 • 0
@7d64535a
Last seen 5 weeks ago
Germany

Hi all,

I have the following type of data:

• 5 batches (5 different data sets)
• 2 conditions (wt, mutant)
• 3 different tissues (data 1 and 2 represent tissue 1, data 3 and 4 represent tissue 2, data 5 represents tissue 3)

I would like to find DEGs in tissue 1 - mutant vs. other tissues (wt and mutant together), but I'm not sure how to specify the parameter for a tissue.

Here is what I would do for only batch and condition for the comparison mutant vs wt:


# preparing coldata
coldata=matrix(,22,3)
coldata[,1]=colnames(countdata)
colnames(coldata)=c("probe","condition","batch")
coldata[1:4,3]=paste("data1")
coldata[5:8,3]=paste("data2")
coldata[9:12,3]=paste("data3")
coldata[13:17,3]=paste("data4")
coldata[18:22,3]=paste("data4")
coldata[c(1:2, 5:6,9:10,13:15,18:20),2]=paste("mutant")
coldata[c(3:4,7:8,11:12,16:17,21:22),2]=paste("wt")

# running DESeq2
dds <- DESeqDataSetFromMatrix(countData = countdata, colData = coldata, design= ~ batch + condition)
dds <- DESeq(dds)
resultsNames(dds) # lists the coefficients
res <- results(dds, name="condition_mutant_vs_wt")


Now I wonder how to include the tissue information and compare tissue 1 - mutant vs. other tissues (wt and mutant together).

DESeq2 • 256 views
1
Entering edit mode
@mikelove
Last seen 1 hour ago
United States

For figuring out the right statistical design to answer questions of interest, I recommend working with a local statistician or someone familiar with linear models in R. Unfortunately, I have to limit my time on the support site to software related questions, and I can't field statistical consultation questions here.

1
Entering edit mode
swbarnes2 ▴ 970
@swbarnes2-14086
Last seen 9 hours ago
San Diego

You can make another column of Coldata whatever you want, and compare subsets of samples to subset of samples that way. But are you sure that you want to compare tissues to each other? Usually those differences are very large.

0
Entering edit mode

Yes, we are aware of large differences between the tissues, but we would still like to do it. I will use then such a design = ~ batch + tissue + condition, thanks a lot for your reply!

For everyone having a similar problem to mine, you can also find these two posts interesting, which together with the answer above clarify a lot: DESEq2 comparison with mulitple cell types under 2 conditions and DESeq2: one condition vs multiple combined