Entering edit mode
9 months ago
pingu77
▴
20
Hello,
I have RNAseq coming from CDX models that were treated with DMSO and another compound (condition). I have 12 CDX models in total, 6 were implanted with cell1 (4 DMSO, 2 treated) and 6 with cell2 (3DMSO, 3 treated).
I would like to run a differential expression analysis using DESeq2 but I am a bit confused about the design to use. I want to pull the 12 samples together and compare DMSO vs treated but I want to take into account that different cell lines were implanted in the mouse.
Would it be correct to use the following design?
dds <- DESeqDataSetFromMatrix(countData = countMatrix,
colData = annotation,
design= ~cell +condition)
Thank you in advance
Please show
annotation
.sure, annotation is a dataframe with the following columns:
Then your suggested design makes sense. It's a paired analysis, adjusting for the cell effect so that the condition effect is tested.
thank you! it's the first time that I adjust for the cell effect and I wanted to be sure! thank you again!
Depending on what cell1 and cell2 are, you might be better off splitting the analysis in two. Just because you can 'correct' for the two different kinds of cells doesn't mean it's the smartest way to proceed.
I would like to run one analysis for 2reasons: 1) I am interested to find the genes that are diff. expressed in both cells as I expect to see the same genes up/down regulated (the two cells are similar in terms of indication and response to the treatment) 2) I do not have 3 treated models for cell1, just 2.. It would be better to have 3 replicates
Why did you delete this question, pingu77?
no idea what happened!