PTM Prophet error crashes run

[Shinya-Watanabe]

I have the same issue with this #1310 unable to resolve. The error happens when searching specific PTM modifications search, like below. I used TMT11 plex. The monoisotopic mass of 156.1151 on CHK and 140.1201 on K are the PTM that I am looking for.

KEEPOLD STATIC EM=1 NIONS=b M:15.9949,n:42.0106,CHK:156.1151,K:140.1201 MINPROB=0.5 MAXTHREADS=1 interact-01MoTrPAC_PASS1B-06_T58_PR_BI_20200224_GY_f15.pep.xml interact-01MoTrPAC_PASS1B-06_T58_PR_BI_20200224_GY_f15.mod.pep.xml

Please wait. PTMProphet is running. It can take a long time.
[INFO:] Using statically set 15 PPM tolerance ... 
[INFO:] Writing file interact-01MoTrPAC_PASS1B-06_T58_PR_BI_20200224_GY_f14.pep.xml.tmp.a31788 ...

[INFO:] Creating 1 threads 
[INFO:] Wait for threads to finish ...
[INFO:] Reading file interact-01MoTrPAC_PASS1B-06_T58_PR_BI_20200224_GY_f14.pep.xml ...
​
[INFO:] processed 1000/12405 spectrum_queries		 ​
[INFO:] processed 2000/12405 spectrum_queries		 ​
[INFO:] processed 3000/12405 spectrum_queries		 ​
[INFO:] processed 4000/12405 spectrum_queries		 ​
[INFO:] processed 5000/12405 spectrum_queries		 ​
[INFO:] processed 6000/12405 spectrum_queries		 ​
[INFO:] processed 7000/12405 spectrum_queries		 ​
[INFO:] processed 8000/12405 spectrum_queries		 ​
[INFO:] processed 9000/12405 spectrum_queries		 [WARNING:] Illegal PTM-shuffled peptide detected with non-matching mass: n[230]ITDFGLC[160]K[653]EGISDGATMK[357]TFC[160]GTPEYLAPEVLEDNDYGR
	Neutral Mass (from pepXML) = 5352.71
	BAD (SpectraST::Peptide) = ITDFGLCK
	Neutral Computed Mass for Evaluation = 895.447
	PPM difference = 832711
[INFO:] please check your specified modification masses for precision or adjust MODPREC= parameter ...
[ERROR:] Scan 30559 ... number of mod combinations doesn't match the number of peptides ... 
[INFO:] Check the precision of specified PTM masses OR Adjust 'MODPREC=' parameter and rerun ... exiting ...
Process 'PTMProphet' finished, exit code: 1
Process returned non-zero exit code, stopping
Processing
 interrupted, stopping PTMProphet

log_2024-09-13_13-28-24.txt

[fcyu]

PTMProphet has some issues when there are fixed and variable modifications on the same site. To make it work, you need to do the following

1. Remove the fixed modifications of 57 on C and 229 on K
2. Remove the variable modification of 42 on protein N-term
3. Add additional variable modifications covering the original fixed modification
4. Add additional variable modifications with the masses equal to the summation of the fixed and variable modifications

The following is the basic idea, but you need to refine it by taking the summation.

variable_mod_01 = 15.9949 M 3
# variable_mod_02 = 42.0106 [^ 1 # remove this one
variable_mod_03 = 57.02146 C 3
variable_mod_04 = 229.16293 K 3
variable_mod_05 = 156.1151+57.02146 C 3
variable_mod_06 = 156.1151+229.16293 K 3
variable_mod_08 = 156.1151 H 3
variable_mod_09 = 140.1201+229.16293 K 3

Keep the fixed modification 229.16293 on Nterm_peptide

And remember to click the "Load Mods from MSFragger Settings" button on the "PTM Site Localization" panel to refresh the PTMProphet commands.

Best,

Fengchao

[Shinya-Watanabe]

Thank you for your prompt response. I changed variable and fixed modifications and updated "PTM Site Localization". I wonder how I should change "Mod tag" in "PTMs" box under Quant (Isobaric).

Best,
Shinya

[fcyu]

This is tricky... I think TMT-Integrator can only analyze one modification at a time....

Best,

Fengchao

[Shinya-Watanabe]

I ran as you suggested and with mod_tag: H(156.1151), C(213.13657), K(385.278), K(369.28302) . Everything worked except TMT-integrator. Is there any way to run only TMT-integrator, so i can run it with the individual mod? Let me know if you have any suggestions.

Also, I have a question about modification setup you provided. I expect when peptide has PTM (156.1151) on C or K, they do not get alkylation or TMT modification. In this case, are the following variable modifications correct?
variable_mod_05 = 156.1151+57.02146 (=213.13657) C 3
variable_mod_06 = 156.1151+229.16293 (=385.278) K 3

Process 'IonQuant' finished, exit code: 0
TmtIntegrator [Work dir: C:\Users\u1440477\Desktop\Motrpac\proteomics\global\MSV000092922\output_heart\S1]
C:\Shared_Files\FragPipe-jre-22.0\jre\bin\java.exe -Xmx107G -cp C:\Shared_Files\FragPipe-jre-22.0\tools\tmt-integrator-5.0.9.jar tmtintegrator.TMTIntegrator C:\Users\u1440477\Desktop\Motrpac\proteomics\global\MSV000092922\output_heart\S1\tmt-integrator-conf.yml C:\Users\u1440477\Desktop\Motrpac\proteomics\global\MSV000092922\output_heart\S1\S1\psm.tsv
TMT-Integrator v5.0.9
TMT-Integrator can't find the corresponding site localization column. Please check if the column: HCK:156.11 is in the psm tables.
Process 'TmtIntegrator' finished, exit code: 1
Process returned non-zero exit code, stopping

~~~~~~~~~~~~~~~~~~~~
Cancelling 1 remaining tasks

log_2024-09-13_15-52-44.txt

[fcyu]

I ran as you suggested and with mod_tag: H(156.1151), C(213.13657), K(385.278), K(369.28302) . Everything worked except TMT-integrator.

TMT-Integrator can only analyze one modification (one mass) at a time.

Is there any way to run only TMT-integrator, so i can run it with the individual mod? Let me know if you have any suggestions.

Yes, you just need to uncheck the steps before TMT-Integrator. Look for the "Run XXX" checkbox on the top left.

Also, I have a question about modification setup you provided. I expect when peptide has PTM (156.1151) on C or K, they
do not get alkylation or TMT modification. In this case, are the following variable modifications correct?
variable_mod_05 = 156.1151+57.02146 (=213.13657) C 3
variable_mod_06 = 156.1151+229.16293 (=385.278) K 3

I "translated" the variable modifications directly from your original settings (57 on C and 229 on K are fixed modifications, and the variable modifications are on the top of them). Please modify it as you see fit.

Best,

Fengchao

[Shinya-Watanabe]

Thank you. I will try to run TMT-integrator as you suggested.

I "translated" the variable modifications directly from your original settings (57 on C and 229 on K are fixed modifications, and the variable modifications are on the top of them). Please modify it as you see fit.

I'm sorry. I modified TMT10 workflow, so it wasn't actually right for my settings. I expect peptide contains either alkylation or 156.1151 mod on Cysteine. If I use the following parameters, would it search peptide only with alkylation or 156.1151 on C but not non-modified or alkylation+156.1151?

fixed_mod = 57.02146 C 
variable_mod = 156.1151 - 57.02146 (= 99.09364) C 3

[fcyu]

Set both fixed and variable modification on C will crash PTMProphet....

Best,

Fengchao

[Shinya-Watanabe]

Oh, that's right. It means there is no way to search peptide only with either one of two modifications but not none or both. If I set each of them as variable modifications, would I get 4 patterns (none, modA, modB, and modA+B)? Is there any way to exclude none and A+B after?

Best,
Shinya

[fcyu]

You can search peptides only with alkylation or 156.1151 on C or non-modified using

variable_mod = 57.02146 C 3
variable_mod = 156.1151 C 3

In this setting, there is no way to not search non-modified peptides.

Best,

Fengchao

[Shinya-Watanabe]

Thank you. It seemed worked.

Back to the first question, I ran ubiquitin proteomics data using TMT10-ubiquitination-K_tmt_plus_ubiq workflow with modification to TMT11. This did not crash PTMprophet although this settings have both variable and fixed modifications on K. It seems like there is a way to search variable/fixed modifications on the same AA residue without crashing PTMProphet.

[fcyu]

Hmm, then I am confused by how PTMProphet deals with the modifications. Maybe you can contact them on https://groups.google.com/g/spctools-discuss

Best,

Fengchao

[Shinya-Watanabe]

I will play around with the fragpipe and if I cannot find a clue I will contact PTMProphet group.

Thank you for your help!
Shinya

[petersclarke]

Hello - I am having an issue within PTM-Sheperd searching for R citrullination as a variable modification within fractionated data. This issue may be related to the above thread involving fixed/variable PTMs attributed to the same site, although I haven't been able to resolve it. I hope you may be able to provide some clarity. Thanks!
`PTM-Shepherd version 2.0.6
(c) 2022 University of Michigan
Using Java 11.0.9.1 on 15360MB memory
Finding spectral data
Indexing data from 129
Fatal error: In dataset "129" could not find mzData for run 20240920_SK_PAD4_60min_60k15k60k_sceHCD_EThcD20_MaxIT100_118_SKBR3_surface_129_500ng_Fxn1_calibrated
Process 'PTMShepherd' finished, exit code: 1
Process returned non-zero exit code, stopping

Cancelling 3 remaining tasks`

[fcyu]

This thread is about PTM Prophet. Could you open a new one?

Thanks,

Fengchao