Philosopher writes incorrect protein info to protein.tsv

[jd690764]

• Describe the issue or question: Hi,

I am importing a dataset into skyline using the pep.xml files following the tutorial and I don't understand why peptides that are not in the combined_modified_peptides.tsv file are imported into the spectral library file in skyline.

During import, I use probability cutoff = 0.95 (the highest prob value for a run in this dataset is 0.9513 for a 1% peptide fdr and all the other ones are much lower - I attached the log file). Out of ~27K peptides in the spectral library ~25K are in the combined_mod_pept file. Some of the extra peptides are coming from peptides assigned to decoy proteins (~700) and a handful others are peptides with 0.95 < prob < 0.9513, but after this I still have >1000 peptides that I am not sure how to explain.

A related question: would it be possible to output a filtered version of the pep.xml files that is consistent with the other result files?

Thank you, Janos

• Upload you log_2022-08-05_01-05-53.txt r log file (If a log file hasn't been generated, go to the 'Run' tab in FragPipe, click 'Export Log', zip the resulting "log_[date_time].txt" file to avoid truncation, then attach the zipped file by drag & drop here.)

[fcyu]

Hi Janos,

Your log file shows that the job crashed due to an error. Can you double check if you sent the correct log file?

Best,

Fengchao

[jd690764]

Oops, sorry about that - here is the correct one. Janos log_2022-08-10_20-00-02.txt s

[fcyu]

Thanks for your log file.

First of all, please upgrade Philosopher to the latest version in case it is due to a bug that has been fixed.

Second, IonQuant complained that your proteins have inconsistency:

2022-08-10 19:59:48 [WARNING] - Protein NP_001019845 has two inconsistent entries in protein.tsv files: NP_001019845,NP_001019845,,125,Homo sapiens,,protein lin-52 homolog,1.000000,0.998500, vs NP_001019845,NP_001019845,,112,Homo sapiens,,protein lin-52 homolog,1.000000,0.999000,NP_001358934.
2022-08-10 19:59:48 [WARNING] - The protein NP_001019845 has different lengths (125 vs 112) in different protein.tsv files
2022-08-10 19:59:48 [WARNING] - Protein NP_001092264 has two inconsistent entries in protein.tsv files: NP_001092264,NP_001092264,,375,Homo sapiens,,FHF complex subunit HOOK interacting protein 1B isoform 2,1.000000,0.999000, vs NP_001092264,NP_001092264,,972,Homo sapiens,,FHF complex subunit HOOK interacting protein 1B isoform 2,1.000000,0.999000,NP_115503.
2022-08-10 19:59:48 [WARNING] - The protein NP_001092264 has different lengths (375 vs 972) in different protein.tsv files
2022-08-10 19:59:48 [WARNING] - Protein NP_001284602 has two inconsistent entries in protein.tsv files: NP_001284602,NP_001284602,,237,Homo sapiens,,regulation of nuclear pre-mRNA domain-containing protein 2 isoform,1.000000,0.999000, vs NP_001284602,NP_001284602,,934,Homo sapiens,,regulation of nuclear pre-mRNA domain-containing protein 2 isoform,1.000000,0.999000,NP_001284603, NP_001374049, NP_001374051, NP_001374052, NP_056018.
2022-08-10 19:59:48 [WARNING] - The protein NP_001284602 has different lengths (237 vs 934) in different protein.tsv files
2022-08-10 19:59:48 [WARNING] - Protein NP_001019845 has two inconsistent entries in protein.tsv files: NP_001019845,NP_001019845,,125,Homo sapiens,,protein lin-52 homolog,1.000000,0.998500, vs NP_001019845,NP_001019845,,112,Homo sapiens,,protein lin-52 homolog,1.000000,0.999000,NP_001358934.
2022-08-10 19:59:48 [WARNING] - The protein NP_001019845 has different lengths (125 vs 112) in different protein.tsv files
2022-08-10 19:59:48 [WARNING] - Protein NP_001092876 has two inconsistent entries in protein.tsv files: NP_001092876,NP_001092876,,433,Homo sapiens,,zinc transporter ZIP6 isoform 2,1.000000,0.999000,NP_036451 vs NP_001092876,NP_001092876,,375,Homo sapiens,,zinc transporter ZIP6 isoform 2,1.000000,0.999000,.
2022-08-10 19:59:48 [WARNING] - The protein NP_001092876 has different lengths (433 vs 375) in different protein.tsv files
2022-08-10 19:59:48 [WARNING] - Protein NP_001092264 has two inconsistent entries in protein.tsv files: NP_001092264,NP_001092264,,375,Homo sapiens,,FHF complex subunit HOOK interacting protein 1B isoform 2,1.000000,0.999000, vs NP_001092264,NP_001092264,,972,Homo sapiens,,FHF complex subunit HOOK interacting protein 1B isoform 2,1.000000,0.999000,NP_115503.
2022-08-10 19:59:48 [WARNING] - The protein NP_001092264 has different lengths (375 vs 972) in different protein.tsv files
2022-08-10 19:59:48 [WARNING] - Protein NP_001284602 has two inconsistent entries in protein.tsv files: NP_001284602,NP_001284602,,237,Homo sapiens,,regulation of nuclear pre-mRNA domain-containing protein 2 isoform,1.000000,0.999000, vs NP_001284602,NP_001284602,,934,Homo sapiens,,regulation of nuclear pre-mRNA domain-containing protein 2 isoform,1.000000,0.999000,NP_001284603, NP_001374049, NP_001374051, NP_001374052, NP_056018.
2022-08-10 19:59:48 [WARNING] - The protein NP_001284602 has different lengths (237 vs 934) in different protein.tsv files
2022-08-10 19:59:49 [WARNING] - Protein NP_001012768 has two inconsistent entries in protein.tsv files: NP_001012768,NP_001012768,,481,Homo sapiens,,abl interactor 1 isoform b,0.999900,0.999000,NP_001012769, NP_001012770, NP_001334959 vs NP_001012768,NP_001012768,,135,Homo sapiens,,abl interactor 1 isoform b,1.000000,0.999000,.
2022-08-10 19:59:49 [WARNING] - The protein NP_001012768 has different lengths (481 vs 135) in different protein.tsv files
2022-08-10 19:59:49 [WARNING] - Protein NP_001037850 has two inconsistent entries in protein.tsv files: NP_001037850,NP_001037850,,408,Homo sapiens,,transmembrane protein 237 isoform a,1.000000,0.999000,NP_689601 vs NP_001037850,NP_001037850,,1212,Homo sapiens,,transmembrane protein 237 isoform a,1.000000,0.998600,NP_001243390.
2022-08-10 19:59:49 [WARNING] - The protein NP_001037850 has different lengths (408 vs 1212) in different protein.tsv files

Did you re-use the old files from the previous run? Does your fasta file have proteins with the same ID but different sequence? If yes, please correct the settings and re-run the whole task from scratch.

Best,

Fengchao

[jd690764]

Hi Fengchao,

I use an older version of philosopher, because: https://github.com/Nesvilab/FragPipe/issues/787.

I saw those messages, but they don't make sense to me - I have been using the same library. I'll re-run it one more time.

Thanks, Janos

[jd690764]

Hi Fengchao,

I re-ran the analysis and I got the same message about some of the proteins having more than one length. I spot -checked the library, that's not correct, so I don't understand why ionquant reports that message. I attached the new log file.

Thanks, Janos log_2022-08-16_10-45-39.txt

[fcyu]

Hi Janos,

Do you mean that those warnings are not correct? Can you send me all of your protein.tsv files?

Thanks,

Fengchao

[jd690764]

What I meant was that in the fasta library there is only one entry for the protein I checked. The protein.tsv files are mixed, 2 out 18 are incorrect, reporting an incorrect protein length.

Janos

[fcyu]

Hi Janos,

Thanks for your clarification. It sounds like a bug in Philosopher.

Hi Felipe @prvst , can you take a look? Basically, Philosopher wrote incorrect entries to the protein.tsv. For example

Protein NP_001014796 has two inconsistent entries in protein.tsv files: 
NP_001014796,NP_001014796,,119,Homo sapiens,,discoidin domain-containing receptor 2 precursor,1.000000,0.998500,NP_001139699
NP_001014796,NP_001014796,,855,Homo sapiens,,discoidin domain-containing receptor 2 precursor,0.997400,0.998600,.

Thanks,

Fengchao

[prvst]

What I meant was that in the fasta library there is only one entry for the protein I checked. The protein.tsv files are mixed, 2 out 18 are incorrect, reporting an incorrect protein length.

Janos

Sorry I don't understand what you're describing, could you elaborate more?

[anesvi]

@jd690764 any update on this issue? Do you still have the problem with philosopher output ?