inferring vs setting MS1, MS2 mass accuracies and scan windows

[zjuezhen]

Hi Vadim,

Thank you very much for developing DIA-NN; we find it quite easy to use and the searches are quite fast.

We performed searches with either:

1. allowing DIA-NN to infer MS1, MS2 mass accuracies and scan windows for each run (report_auto_params.log.txt) or
2. setting MS1, MS2 mass accuracies and scan window to the "Averaged recommended settings for this experiment" from 1) (report_set_params.log.txt)

and found slightly more IDs from 2). What could be the reason for this discrepancy?

Thank you and best regards,

Jenny

[vdemichev]

Hi Jenny,

It's a random effect basically. You can play with mass acc values manually to see which produce best results in your case.

Best, Vadim