DIA-NN is running 20k+ minutes to create a speclib of rattus novergicus proteome

[ThalisNovo]

Hi Vadim,
I ran and analysis to myt phd and the SpecLib generation is running for more than 20k minutes and taking 98% of my 64gb RAM without succed. I ran proteomes for fungi with 8k proteins and they got complete, but without the PDF report (I dont know if they are parts of the same issue).

My computer has:
24 cores (i9 12gen) with 32gb RAM (reaches 64gb due to dual channel);
a 2k SSD NvMe
a geForce 4060Ti 8gb

My data are from a Synapt QToF from Waters in nanoUPLC in DIA in 5 fractions and 3 replicates from control and treated.
I converted the 30 fractions (control and treated in 3 replicates and 5 fractions each) in MzML using the GNPS waters converter and then put them on DIA-NN with the following parameters.

I've been learning how to use DIA-NN the past two months by myself with the documentation available. Any tips or extra direction would be very helpful.

Thanks Vadim for the attention

[ThalisNovo]

I found out the log for this run in rattus novergicus. I will attach below, it might help to understand the problem.

Spectral_Library_Rattus.log.txt

[vdemichev]

Hi Thalis,

The log attached is just for library generation, not the analysis.

To reduce RAM usage, please

• Switch to DIA-NN 1.9.2
• Limit precursor charge range to 2-3
• Try first without variable modifications

Best,
Vadim

[ThalisNovo]

Thanks Vadim for the response, Im reanalyzing the data with your suggestions. May I and my supervisor contact you by email? We want to discuss about this analysis and a possible collab :)