Hi, I have a count matrix with 8 samples (3 replicates each) and I performed a DESeq2 analysis. I did a PCA analysis and there was quite nice clustering of the replicates of the 8 samples separately. I want to show the PCA analysis for 2 to 3 samples together, and I wanted to see how I could go about doing this. Would I need to make a separate count matrix for the comparison of the different samples or is there a way I can do this by using the same count matrix but specifying a different data matrix for DESeq2? Sorry I'm new to DESeq2 so this might be a very basic question. 

Thanks so much!

I don’t follow. Why would you want a PCA analysis with only 2 or 3 samples. It’s much more useful to see the relationship of all 8. I wouldn’t recommend subsetting to only show 2 or 3.