Cutoff value for shrunken logFC

Question

[official.profile]

Hello

Is there any rule of thumb for filtering shrunken logFC, similar to |logFC|>1.5 for raw results? Or perhaps one should find the cutoff purely based on the distribution of shrunken logFC? I think some standardized value is needed though, e.g. from the perspective of volcano plots.

Kind regards

[Shred]

I think this question was answered so many times here. You need to determine a threshold depending on experimental setup, type of experiment, biological questions.. shrunken logFCs have nothing different (in term of how to manage them) than non-shrunken logFCs.

[official.profile]

I tried to find the answer on the forum but couldn't find one. Please be more specific.

Answer 1

[ATpoint]

There is no rule for this, nor is 1.5 a good choice beyond that many papers do it. Entirely depends on you. If you have lots of DEGs and need to prioritize them then maybe take 2 or 3, if you barely have any then take everything that is significant regardless of effect size (imo).

[official.profile]

Thank you for your reply :)