RNA-seq for differential gene expression analysis with only one biological replicates

Question

[WMA]

Hello everyone,

I am working on my master's thesis and I am facing huge problems I cannot fix due to time and money problems.

I sent six samples of RNA to a sequencing company, 3 treated and 3 controls. And only two from each have passed their quality test. I am only testing one drug by the way.

I did differential gene expression analysis and got the results. I am only focusing on significant pathways, I will not focus on genes. From what I read genes are out of the question at this point.

My advisor says everything is ok but I think my thesis is a joke. Will I fail to defend my thesis?

I also did three apoptosis assays: Nuclear morphology by Hoechst 33324 staining method, acridine orange–ethidium bromide dual staining method, and Annexin-V/ propidium iodide apoptosis assay in flow cytometry.

Thank you tons, I asked this question before but without these details that might give me a different answer.

Answer 1

[ATpoint]

And only two from each have passed their quality test

2 vs 2 is not very powerful but it's ok to proceed. Common DE tools like limma will work.

Will I fail to defend my thesis?

Cannot answer. No details provided and also off-topic here. Finish the thesis best possible, then move on. Student thesis very often have shortcomings, often due to poor planning or guidance from the supervisors, but that's nothing you can change right now, nor is it worth losing sleep over. Learn from mistakes and do better in your PhD or job once finished.

[WMA]

Thank you.

Answer 2

[matthew.pace]

I was in a similar situation for my masters (arguably worse, I only had a single sample per treatment). It's unlikely that your work would be published in any high-impact journals, but it is sufficient for a masters thesis. Just make sure to make it crystal clear that you understand the shortcomings of having too few samples and explain that you were limited financially. I'm sure your examiners will find this understandable. Good luck!

[WMA]

Thank you!!

[dsull]

2 samples per replicate is more than sufficient; more is always better but two is perfectly fine. You might be missing some signal that you would've gotten with 3 samples, but your analysis will still be fine.

[WMA]

Thank you. My supervisor just told me something I cannot believe. He said since the treatment replicate are very similar he will only use one! He says it will cost money to do two and it is unnecessary! I am dead inside

[WMA]

Matthew I hope you comment on this because now I became like your scenario!

Answer 3

[yhdist]

I don't believe this will make you fail your thesis. Just remember to either point this out in your writing, or to have your answer ready if the question does come up during your oral presentation. (your supervisor will most likely suggest the latter, but it depends on where you are from)

[WMA]

Thank you very much but my supervisor just throw a bomb at me, he says since the treatment replicate are very similar he will only use one! He says it will cost money to do two and it is unnecessary!