PCA plot

Question

[Aaliya]

Hello everyone, I had posted this question before: PCA plot (Suggestions Needed)

I did my PCA again but on only one study (three control and three treatment conditions), and this is my result:

I have redone the codes, and this is the result I am getting. I wanted your opinion on if it's a good result or not.

I have 8 such studies (I am unable to get the PCA). Is there any other statistical tool that can take in my 4 WT studies and 4 Mutant studies? Each study has a DESeq2 output generated.

Thank you!

[ATpoint]

What is your question? Add details and clarity. Ask a precise question and please avoid opening more questions on the same underlying issue. What is the analysis goal? As I said before, PC1 drives condition differences and control is more variable than treatment.

[jkim]

cross posted in bioconductor.

https://support.bioconductor.org/p/9157968/

[LauferVA]

Without more context, we cannot interpret this PCA plot. There are many possibilities. In addition, this question is basic enough that it would actually be irresponsible of me to try to answer it. I am not trying to be mean, this is just me really trying to tell you the truth and help you. It's just that there are many kinds of quality control methods for RNA-seq, and they all tell you something different, and you really won't know about your data till you have looked at several of them thoughtfully.

The question above implies it is likely there are many things you need to understand, and I cant type it all out. So, its better for me to give a different kind of advice. Here are some ideas:

• If you are affiliated with a university and have a mentor, ask your mentor to help you to find a biostatistician to collaborate with you.
• If you are in a stats class, ask your professor if they could help you with this
• If neither of those, try going to a statistics help desk
• If not affiliated with a university at all, go to DESeq2 vignettes, then open up pubmed, wikipedia, and chatGPT. Go through each kind of quality control measure in the DESeq2 vignettes, then read about them in the other 3. Read and cross-check the information until you have a good understanding of: 1) why each step is taken 2) how each step works ... and so on.

Good luck.