search

vdemichev / DiaNN

DIA-NN - a universal automated software suite for DIA proteomics data analysis.
Other
281 stars 54 forks source link

Issue with thermo .raw files in DIA-NN #916

Closed tannishthadas closed 9 months ago

tannishthadas commented 9 months ago

I am loading thermo .raw files onto the DIA-NN input box but it is giving this error message :

[0:15] File #1/1 [0:15] Loading run C:\users\user\Downloads\19012024.raw No MS2 spectra: aborting ERROR: cannot load the file, skipping [0:15] 0 library precursors are potentially detectable [0:15] Processing... [0:15] Removing low confidence identifications [0:15] Searching PTM decoys [0:15] Removing interfering precursors [0:15] Too few confident identifications, neural networks will not be used [0:15] Number of IDs at 0.01 FDR: 0 [0:15] Calculating protein q-values [0:15] Number of proteins identified at 1% FDR: 0 (precursor-level), 0 (protein-level) (inference performed using proteotypic peptides only) [0:15] Quantification [0:15] Quantification information saved to C:\users\user\Downloads\19012024_PE_UT.raw.quant.

But I have MS2 spectra. This data was acquired from the Orbitrap Fusion Tribrid MS machine. Please help!

vdemichev commented 9 months ago

Please install MSFileReader using the link here https://github.com/vdemichev/DiaNN?tab=readme-ov-file#raw-data-formats

tannishthadas commented 9 months ago

Thanks. i think that worked. but now i am getting stuck like this -

[0:17] File #1/1 [0:17] Loading run C:\users\user\Downloads\19012024.raw [0:29] 2387263 library precursors are potentially detectable [0:29] Processing... [54:01] Cannot perform mass calibration, too few confidently identified precursors [54:01] Cannot perform MS1 mass calibration, too few confidently identified precursors Could you please help me out?

vdemichev commented 9 months ago

It could be that nothing is identifiable - i.e. failed run for some reason. A quick check with a public spectral library (a library is more sensitive than lib-free search on a single run) will make sense.

tannishthadas commented 9 months ago

Thank you so much.

On Sun, Feb 4, 2024 at 5:01 PM Vadim Demichev @.***> wrote:

It could be that nothing is identifiable - i.e. failed run for some reason. A quick check with a public spectral library (a library is more sensitive than lib-free search on a single run) will make sense.

— Reply to this email directly, view it on GitHub https://github.com/vdemichev/DiaNN/issues/916#issuecomment-1925714355, or unsubscribe https://github.com/notifications/unsubscribe-auth/BFTDIDKL6DPPLINRB3LK42TYR5WSHAVCNFSM6AAAAABCKGU75OVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMRVG4YTIMZVGU . You are receiving this because you authored the thread.Message ID: @.***>