Differential-expression analysis using logistic regression in Seurat - glm.fit warning message

Question

[Lucy]

Hi,

I am performing differential-expression analysis in Seurat using the logistic regression test with "donor" as a latent variable. While it does give me an output, I am getting the following warning message.

Warning message:
glm.fit: fitted probabilities numerically 0 or 1 occurred

What does this mean and should I be concerned about it?

Many thanks,

Lucy

Answer 1

[jared.andrews07]

This crossvalidated question answers this quite well. In short, this is likely due to certain genes having no counts in certain donors leading to perfect separation for that variable (which in most cases is odd to see in logistic regression, hence the warning) and can (probably) be ignored safely.

[Lucy]

Ah great, thank you.