Can you use salmon bootstraps to increase sample size for machine learning

Question

[yryan]

Hi folks,

I've got some time course matched RNA seq experiments with n=6 for controls and exposures. I'm trying to do some machine learning classification with these. I'm wondering if using the bootstraps generated by salmon is a valid way of increasing my sample size or is this not really suitable?

Answer 1

[ATpoint]

No, it is not. These are uncertainty estimates, not replicates.

[yryan]

Okay, thanks! (making up characters to comment)