Comet returning PSM with unexpected m/z PPM differences

[Erik Johnson]

Hello hello Comet and crux maintainers,

Thanks in advance for your help and happy new year!

In my parameter file, I set peptide_mass_tolerance = 20.0 and peptide_mass_units = 2. So I expect to see target PSMs where the peptide m/z and spectrum precursor m/z differ by at most 20 PPM. But I'm seeing some PSMs with very large PPM differences.

There seems to be a limit on the length of a message so I'll send the details in a follow-up message in the thread....

[Erik Johnson]

Here's an example: when I run Comet on a particulary spectrum, 762744, which has precursor m/z 334.67004055447 and precursor charge 2

crux comet --parameter-file ./crux.comet.params \

--output-dir ./ \

--overwrite T \

./762744.mzML ./proteome.fasta

(the files are attached to this email), the top target PSM for scan 762744 is the peptide KIPVSSAP which at charge=2 has m/z = 399.73962846688 which differs from the spectrum's precursor m/z by ~2491 PPM. 

Am I missing something? Thanks for your help and thanks for crux!

Best,

Erik Johnson

Postdoc @ CU Boulder

[Erik Johnson]

Looks like there's some limit to the size of an email. I'm getting this error trying to attach the files:

So here's a Drive where you can find the files: https://drive.google.com/drive/folders/1CrDPE2yvauNpcwNK-kyxChBkTZvjutih?usp=drive_link

- Erik 

[jke...@gmail.com]

Can you make the files crux.comet.params and 762744.mzML available or send them to me directly?  Hopefully whatever is wrong is easy to identify given this is such a basic issue.  

Jimmy

[jke...@gmail.com]

Erik, thanks for passing along the relevant files.  Using my tools to read your mzML, I'm seeing the precursor m/z as 400.7379.  That explains why Comet is returning the peptide KIPVSSAP with neutral mass 797.4647 that you're seeing (because of this m/z and the isotope_offset setting).  Do you know where the 334.6700 m/z came from?

------- Scan 762744 header information -------
     filterLine:
       nativeId: scanId=762744
        msLevel: 2
  precursorScan: 0
    precursorMZ: 400.7379
     precursorZ: 2
     activation: HCD
       numPeaks: 105
       scanTime: 762.74
          lowMZ: 62.36
         highMZ: 1666.50

Here's the relevant m/z encoded in the mzML:

[Erik Johnson]

Hey Jimmy,

Thanks for the help. I don't know where 334.6700 m/z came from. Must have been an end of the day, tired brain error. Sorry about that. I'm also seeing a 400.7379 m/z (screenshot below) but that still differs from the peptide m/z by ~2491 PPM. I wasn't familiar with the isotope_error parameter so it sounds like that's probably where my confusion lies. I'll do a bit of reading on that. Thanks for your help.

- Erik

On Tue, Jan 20, 2026 at 5:24 PM Jimmy Eng <jke...@gmail.com> wrote:

fwiw, if you are running Crux-Comet under Linux, I've noticed that it's ~60% slower than running standalone Comet under Linux.  Windows doesn't have the same issues.  I'm not involved directly with the Comet integration in Crux so I have no clue why it's slower under Linux vs. the standalone Comet code that I work on.

On Tue, Jan 20, 2026 at 3:54 PM Jimmy Eng <jke...@gmail.com> wrote:

I've got the files.  Hopefully I can figure out what's causing the mass error issue you're seeing.

On Tue, Jan 20, 2026 at 3:47 PM Erik Johnson <airik.k...@gmail.com> wrote:

Hey Jimmy,

Wow you're fast! thank you. I put the files here: https://drive.google.com/drive/folders/1CrDPE2yvauNpcwNK-kyxChBkTZvjutih?usp=drive_link. I'll try attaching to this email too...

--
You received this message because you are subscribed to a topic in the Google Groups "crux-users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/crux-users/nJQ07nCN20o/unsubscribe.
To unsubscribe from this group and all its topics, send an email to crux-users+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/crux-users/d6df25ef-ff8a-46c2-91da-cc307d7818ean%40googlegroups.com.

[jke...@gmail.com]

Erik,

Here's an explanation of the "isotope_error" parameter in Comet.  Your search was performed with that parameter set to "2" but if you prefer to not use it, which would simplify your ppm error analysis, simply set that parameter to "0".  Otherwise, to calculate ppm error with the existing search results, you might consider calculating the ppm error using zero, one, and two 13C offsets (1.00335483) and take the minimum error of the three.