Different parameters result in different results

[mcburke-msdc]

• The same raw data has been searched using semi-selective and non-selective enzyme cleavage rules. There are multiple cases where the same scan number was assigned to the same sequence in both searches; however, the num_matched_ions and hyperscore are higher for the non-selective search. I am a bit puzzled by this as the m/s values in the query and the theoretical fragment ions are the same for both cases. I am happy to share any data files. Thank you in advance.

[fcyu]

I guess the parameters such as use top N peaks and min ratio are different. Could you share both log files?

Thanks,

Fengchao

[mcburke-msdc]

Thank you for your assistance. I've attached the log files from the selective ("Sel_FragPipe0.log") and non-selective ("NonSel_FragPipe0.log") searches.

Please let me know if you have any other questions.

Appreciatively, Meghan

From: Fengchao @.> Sent: Tuesday, November 7, 2023 9:56 AM To: Nesvilab/FragPipe @.> Cc: Burke Harris, Meghan C. (Fed) @.>; Author @.> Subject: Re: [Nesvilab/FragPipe] Fragmentation rules for selective vs non-selective enzyme cleavage? (Issue #1322)

I guess the parameters such as use top N peaks and min ratio are different. Could you share both log files?

Thanks,

Fengchao

- Reply to this email directly, view it on GitHubhttps://github.com/Nesvilab/FragPipe/issues/1322#issuecomment-1798771972, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AVHW6MQILMVHZLOQ2MWHVH3YDJDXTAVCNFSM6AAAAAA7BKJ63WVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOJYG43TCOJXGI. You are receiving this because you authored the thread.Message ID: @.**@.>>

[fcyu]

I couldn't find the log files. GitHub truncated them. Please go to https://github.com/Nesvilab/FragPipe/issues/1322 and upload them there.

Best,

Fengchao

[mcburke-msdc]

Uploading the log files again. Please let me know if you have any trouble.

NonSel_FragPipe0.log Sel_FragPipe0.log

[fcyu]

Just as what I guessed, the following are the parameters that made the differences:

Best,

Fengchao

[dpolasky]

If you're wondering why the parameters are different when they weren't set differently, it's because of the parameter optimization done in MSFragger - it tests several parameters and chooses the ones that give the most IDs in a restricted "first search". With the different cleavage settings, different parameters ended up getting chosen - hence the different results. You can disable the parameter optimization to guarantee the same parameters for doing comparisons.

Best, Dan

[mcburke-msdc]

Thank you for your prompt responses, Fengchao and Dan. It is, indeed, the parameter optimization that caused the different settings. May I ask if it's possible to write to the output file exactly which ions were matched and counted in the "num_matched_ions" field?

[fcyu]

We can let MSFragger "re-match" the ions after the database searching (just like what we do for the spectral library building), but it might not be exactly the same as those used in the database searching since it is infeasible to keep track of the ions used in the database search.

Best,

Fengchao