search

vdemichev / DiaNN

DIA-NN - a universal automated software suite for DIA proteomics data analysis.
Other
283 stars 53 forks source link

How to Populate the 'report-first-pass.phosphosites_90' file #1129

Open martingolkowski opened 3 months ago

martingolkowski commented 3 months ago

Hi, thanks for publishing this tool, I really love it!

There is one thing that came up when I switched from DIA-NN 1.8.1 to 1.9.1 recently:

When computing dia phosphpopeptide data using the DIA-NN generated spectral library, I saw that DIA-NN creates an output file named 'report.phosphosites_90' file. This file is incredibly useful for downstream data analysis, as it contains the information on the exact location of the phosphosite in the protein sequence, including the +/- aa window. However, this file does not contain many of the phosphopeptides that were identified on first pass. For that, there seems to be an output file named 'report-first-pass.phosphosites_90', but I cant get it to populate. Is there any setting I need to enable running DIA-NN to populate the 'report-first-pass.phosphosites_90' file?

Thank you so much for you support in advance!

Cheers,

Martin

vdemichev commented 3 months ago

Hi Martin,

The phospho matrices are filtered in a fairly stringent way to exclude spurious false IDs. I strongly recommend to never use first pass data when doing lib-free phospho, it's really not going to work well. However you can always reanalyse the data without MBR (just check 'Use existing quant files' to make this reanalysis quick) - this will lead to those matrices being produced by DIA-NN effectively for the first pass (but the results will be much worse than with MBR).

Best, Vadim

martingolkowski commented 3 months ago

Thanks Vadim for the clarification, this is super helpful!

Cheers,

Martin

From: Vadim Demichev @.> Sent: Tuesday, August 13, 2024 9:05 AM To: vdemichev/DiaNN @.> Cc: MARTIN GEORG GOLKOWSKI @.>; Author @.> Subject: Re: [vdemichev/DiaNN] How to Populate the 'report-first-pass.phosphosites_90' file (Issue #1129)

Hi Martin,

The phospho matrices are filtered in a fairly stringent way to exclude spurious false IDs. I strongly recommend to never use first pass data when doing lib-free phospho, it's really not going to work well. However you can always reanalyse the data without MBR (just check 'Use existing quant files' to make this reanalysis quick) - this will lead to those matrices being produced by DIA-NN effectively for the first pass (but the results will be much worse than with MBR).

Best, Vadim

— Reply to this email directly, view it on GitHubhttps://github.com/vdemichev/DiaNN/issues/1129#issuecomment-2286482175, or unsubscribehttps://github.com/notifications/unsubscribe-auth/BKPOC5F77QGM4XOX7CIW42LZRIOCFAVCNFSM6AAAAABMMLRDDWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEOBWGQ4DEMJXGU. You are receiving this because you authored the thread.Message ID: @.***>