Nesvilab / FragPipe

A cross-platform Graphical User Interface (GUI) for running MSFragger and Philosopher - powered pipeline for comprehensive analysis of shotgun proteomics data
http://fragpipe.nesvilab.org
Other
181 stars 37 forks source link

TMT-integrator feature-request/discussion: summarize/group intensities by modified sequence #801

Open MiguelCos opened 2 years ago

MiguelCos commented 2 years ago

Dear Fragpipe team,

As always, many thanks for the amazing work you do developing maintaining Fragpipe and all its tools.

Would it be possible to offer intensity grouping by modified sequence?

We have noticed that, when grouping by peptide, we are not able to quantitatively differentiate between different modification versions of the same peptide sequence after processing with TMT-integrator.

I would illustrate why we think this is essencial with an example of our applications:

We are interested in N-terminomics/analysis of proteolytic processing, and it crucial for us to differentiate between N-terminally acetylated peptides vs N-terminally-TMT-tagged peptides. This is not possible with the current sequence-focused approach.

I know that it is possible to summarize by PTM site, but I believe that we then lose the information for non-modified peptides; please correct me if this is not right.

I would really appreciate your feedback on this front.

Best wishes, Miguel

anesvi commented 2 years ago

Hi Miguel,

You would need to discuss with Hui-yin, who is the lead TMT-Integrator developer, if she can add another Index (peptide+modification) to TMT-Integrator.

The problem is, do you need to distinguish just some modifications, or all? e.g. PNSTM[+16]EMK PNSTMEM[+16]K PNSTM[+16]EM[+16]K Acetyl-PNSTM[+16]EMK Acetyl-PNSTMEM[+16]K …

So many possibilities to group (e.g. ignoring common mods like M+16 or not). Also for TMT, what would one do with TMT midications (there could be partial labeling and full labeling)

So it gets complicated as a general cases. Perhaps something can be done specifically for termiomics data.

Best Alexey

From: Miguel Cosenza-Contreras @.> Sent: Wednesday, August 17, 2022 11:02 AM To: Nesvilab/FragPipe @.> Cc: Subscribed @.***> Subject: [Nesvilab/FragPipe] TMT-integrator feature-request/discussion: summarize/group intensities by modified sequence (Issue #801)

External Email - Use Caution

Dear Fragpipe team,

As always, many thanks for the amazing work you do developing maintaining Fragpipe and all its tools.

Would it be possible to offer intensity grouping by modified sequence?

We have noticed that, when grouping by peptide, we are not able to quantitatively differentiate between different modification versions of the same peptide sequence after processing with TMT-integrator.

I would illustrate why we think this is essencial with an example of our applications:

We are interested in N-terminomics/analysis of proteolytic processing, and it crucial for us to differentiate between N-terminally acetylated peptides vs N-terminally-TMT-tagged peptides. This is not possible with the current sequence-focused approach.

I know that it is possible to summarize by PTM site, but I believe that we then lose the information for non-modified peptides; please correct me if this is not right.

I would really appreciate your feedback on this front.

Best wishes, Miguel

— Reply to this email directly, view it on GitHubhttps://github.com/Nesvilab/FragPipe/issues/801, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AIIMM63I3PKWWZJ4KPP4ERLVZT5FXANCNFSM56Z4RM6A. You are receiving this because you are subscribed to this thread.Message ID: @.**@.>>


Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues

MiguelCos commented 2 years ago

Hello Alexey,

Many thanks for your answer and your input.

It is true that it is not necessary to differentiate between every modification, but some are particularly interesting biologically.

In terms of N-terminomics + TMT-experiments: N-terminal Acetylation vs N-terminal TMT help us know if a truncation comes from potential proteolytic processing or something like a shifted translation initiation site.

I am working on some attempts to summarize the psm.tsv based on modified sequences in this repo, and it seems to work well for single mixture experiments, but it would get more tricky for me for multi-mixture experiments.

Maybe Hui-yin would have an idea of how complicated it would be to implement some kind of selective peptide+modification based summarization (i.e. defining interesting modifications), and/or if it is worth the effort.

Best wishes, Miguel

huiyinc commented 2 years ago

Hi Miguel,

Sorry for the late reply. It took me some time thinking how to properly define a new index that fits your needs. Based on your description, I think you would like to compare peptide sequences with and without a specified modification at the same time, right? So, the new index should be able to indicate peptides even if they don't have the modification. Another question I have is how to distinguish modified peptides. For example, a peptide, VETGVLKPGMVVTFAPVNVTTEVK, is assigned with two different modifications (as listed below), would you consider them as the same or different modified peptides?

01CPTAC_CCRCC_P_JHU_20171106_LUMOS_f01.41377.41377.4 | VETGVLKPGMVVTFAPVNVTTEVK | 10M(15.9949), 24K(229.1629), 7K(229.1629), N-term(229.1629) 01CPTAC_CCRCC_P_JHU_20171106_LUMOS_f01.45754.45754.4 | VETGVLKPGMVVTFAPVNVTTEVK | 24K(229.1629), 7K(229.1629), N-term(229.1629)

Maybe we can have this discussion in private? My email address is: @.*** Thanks.

Huiyin

Miguel Cosenza-Contreras @.***> 於 2022年8月19日 週五 下午5:27寫道:

Hello Alexey,

Many thanks for your answer and your input.

It is true that it is not necessary to differentiate between every modification, but some are particularly interesting biologically.

In terms of N-terminomics + TMT-experiments: N-terminal Acetylation vs N-terminal TMT help us know if a truncation comes from potential proteolytic processing or something like a shifted translation initiation site.

I am working on some attempts to summarize the psm.tsv based on modified sequences in this repo https://github.com/MiguelCos/summarize_psm_tsv_fragpipe, and it seems to work well for single mixture experiments, but it would get more tricky for me for multi-mixture experiments.

Maybe Hui-yin would have an idea of how complicated it would be to implement some kind of selective peptide+modification based summarization (i.e. defining interesting modifications), and/or if it is worth the effort.

Best wishes, Miguel

— Reply to this email directly, view it on GitHub https://github.com/Nesvilab/FragPipe/issues/801#issuecomment-1220457733, or unsubscribe https://github.com/notifications/unsubscribe-auth/ALAWWA35MSH5Z5CAZJT6LK3VZ5HRXANCNFSM56Z4RM6A . You are receiving this because you are subscribed to this thread.Message ID: @.***>

-- Hui-Yin Chang, 張彙音 Assistant Professor Department of Biomedical Sciences and Engineering National Central University, Taiwan

MiguelCos commented 1 year ago

Dear Huiyin (@huiyinc),

I am very sorry for the long response time.

I am answering your questions now:

Yes, I would like to be able to distiguish the same peptide with different modifications, or no modifications at all.

Something like the combined_modified_peptide.tsv that is generated after IonQuant quantitation.

We are particularly interested in differences in the N-term or C-term, because we regularly perfom semi-specific searches to look for intrinsic proteolytic activity, and (for example) Nterm-acetyl vs Nterm-TMT help us define a peptide as potential product of endogenous proteolysis or not.

In your two exemplary PSMs, I would summarize those together in the context of our biological interest, because the variable oxidation of Met does not change anything in biological terms.

I can't see your email address from the GitHub discussion (it is "censored" in the post), but I would really like to get back to this discussion if you are able to.

If you have time, you can contact me via my institutional email:

miguel.de.jesus.cosenza.contreras@uniklinik-freiburg.de

It would be wonderful if we can talk about a potential implementation of modified peptide summarization in TMT-integrator.

Best wishes, Miguel