Creating DESeq object and choosing a contrast
1
0
Entering edit mode
Dev Raj • 0
@70c54c57
Last seen 18 days ago
Nepal

I am new to this RNAseq world

I want to perform differential gene expression analysis using DESeq2. I read the DESeq2 vignette but got confused in the Contrasts and Interaction section.

My metadata consists of 22 samples of adipose tissue, 12 from high diet-fed mice (HFD) and 10 from chow-fed mice (CD) each co-cultured with tumor (T) and control (C).

I have seen that adipose tissue from HFD already differs significantly from adipose tissue from CD (without tumor) but I want to test and filter out only those differences that resulted due to tumor co-culture in HFD compared to CD.

I used the following code to create the DESeq object

dds <- DESeqDataSetFromMatrix(countData = main_data, colData = meta_data, design = ~treatment + tumor_status + treatment:tumor_status)
dds <- DESeq(dds)
resultsNames(dds)

[1] "Intercept"                  "treatment_HFD_vs_CD"        "tumor_status_T_vs_C"        "treatmentHFD.tumor_statusT"


Is my approach of creating the DESeq object correct ? Also which contrast will give me the desired result?

DESeq2 • 109 views
0
Entering edit mode
@mikelove
Last seen 1 day ago
United States

If you've read over the section on interactions and are still confused what these terms mean, I'd recommend consulting with a statistician or someone familiar with linear models in R. Picking and interpreting the results table is a key part of the experiment.