Deseq2 returned the same p-adj for several genes

Question

[Matt]

I'm working with cells under treated and untreated conditions. Deseq2 gave me these results. No p-adj bellow 0.05 and the same value for several genes. Are the results valid? Should I try something else?

Thank you for your attention!!

[Ram]

Did you do a basic google search before creating a new post here? There are a lot of posts on multiple forums addressing this question.

https://www.google.com/search?q=deseq2+padj+all+the+same

[Matt]

I'm sorry. I didn't mean to cause you any inconvenience, Ram. I'm old and not sure how the Web works yet. This blog thing is new territory for me. Again, I'm sincerely sorry if my question upset you, Ram. It was not my intention. My colleague created this profile a while ago, in case I needed some help. Thank you for your reply. It was very helpful. Best regards.

[Ram]

You're not causing me any inconvenience, this is not personal. It is now widely known that using Google for any question you have serves up possible solutions or at least connects you to other people that have had a similar experience or have similar questions. People abuse this availability of information to think they're qualified to question experts (like being skeptical of all vaccines because some blog mentioned some lone study) but not using this ubiquitous information to get quick answers to your question is just weird. Any way, that's one way the Web works in my opinion - you use a search engine to get to the information you need.

Answer 1

[Kevin Blighe]

In a nutshell: your null hypothesis is invalid. Please re-think your experimental design.

Kevin

---

NB: Kevin Blighe I'm restoring the content. Please do not delete it again.

[Ram]

Is the null hypothesis invalid or is there no evidence against the null hypothesis (like what dariober says here: same padj for all the genes after DEseq analysis)

[Kevin Blighe]

Either / or, Sir.

[Ram]

They sound like opposites - am I missing something?

[Kevin Blighe]

Who cares. It won't solve World hunger.

[Ram]

On a serious note, what is an "invalid" null hypothesis? Does a null hypothesis not just mean absence of an explanation? How can one have an invalid absence when there's nothing to actually be valid or invalid?

[Kevin Blighe]

You are raising philosophical questions that are causing my brain to explode ¯\_(:/)_/¯ I believe, with my answer, I was implying that the hypothesis that gene expression differs between the conditions being compared was invalid, and that one needed to re-think this set / design.

[Ram]

Ah, I see "the hypothesis that gene expression differs between the conditions" I think that's the alternative hypothesis though, the null is "there is no difference between the conditions". Am I mistaken?

[Kevin Blighe]

Could be, Sir Ram.