More than one glycosylation site

[ValentinaRangelAngarita]

I've run a couple of searches with MSFragger - Glyco, and I have noticed that it only identifies one glycosylation site (despide increasing the number of possible modifications). Is there any way to change this? I know my sample should have 1-3 glyco sites based on other searches done in other programs.
Thanks!

[fcyu]

If you are using the glyco workflow, MSFragger-glyco can only detect one glycosylation per peptide, unless you change the mass offsets list to contain the masses equal the summation of multiple glycosylations. In that case, you MIGHT be able to see some peptides with multiple glycosylation, but you will have to do the "deconvolution" by yourself because MSFragger-Glyco only reports the total masses of the glycosylations.

Best,

Fengchao

[dpolasky]

Are you looking for multiple glyco sites in a single peptide, or on a whole protein? As Fengchao said, we currently can only identify a single mass shift per peptide (which can be the summed mass of glycans on multiple sites), but we cannot currently break down the mass into individual sites, though we are working on adding a way to do this in the future.

[ValentinaRangelAngarita]

Thank you for the quick response! I am asking for a single peptide - I had run the search with O-Pair and Byonic before, and I know a couple of peptides that have more than one glycosite. Thanks for letting me know!

…

On Tue, Jun 15, 2021 at 11:24 AM Daniel Polasky ***@***.***> wrote: Are you looking for multiple glyco sites in a single peptide, or on a whole protein? As Fengchao said, we currently can only identify a single mass shift per peptide (which can be the summed mass of glycans on multiple sites), but we cannot currently break down the mass into individual sites, though we are working on adding a way to do this in the future. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#160 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AUPYDK3NTG2NGZEFBZ53LP3TS5WEVANCNFSM46XPKR6A> .

-- Valentina Rangel Angarita (They/She) PhD Student - Malaker Lab https://www.malakerlab.com/ Department of Chemistry Yale University

[anesvi]

With N-liked, the changes of getting 2 sites (with two NxS/T motifs) on the same peptide are quite low. Are you doing O-linked glyco? Yes, we are currently assigning it as a single mass shift on one residue, even though some may actually be 2 different glycans. But we do not have that functionality like in O-pair yet. Alexey From: ValentinaRangelAngarita ***@***.***> Sent: Tuesday, June 15, 2021 11:34 AM To: Nesvilab/MSFragger ***@***.***> Cc: Subscribed ***@***.***> Subject: Re: [Nesvilab/MSFragger] More than one glycosylation site (#160) External Email - Use Caution Thank you for the quick response! I am asking for a single peptide - I had run the search with O-Pair and Byonic before, and I know a couple of peptides that have more than one glycosite. Thanks for letting me know!

…

On Tue, Jun 15, 2021 at 11:24 AM Daniel Polasky ***@***.***<mailto:***@***.***>> wrote: Are you looking for multiple glyco sites in a single peptide, or on a whole protein? As Fengchao said, we currently can only identify a single mass shift per peptide (which can be the summed mass of glycans on multiple sites), but we cannot currently break down the mass into individual sites, though we are working on adding a way to do this in the future. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#160 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AUPYDK3NTG2NGZEFBZ53LP3TS5WEVANCNFSM46XPKR6A> .

-- Valentina Rangel Angarita (They/She) PhD Student - Malaker Lab https://www.malakerlab.com/ Department of Chemistry Yale University — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#160 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AIIMM6Y5OFZ2OGTPNE4L7BLTS5XFBANCNFSM46XPKR6A>. ********************************************************** Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues

[HansABakker]

Hello,
I was on the way to install MSFragger when I stumbled over this. It is a few years old and you mention you are working on this.
This is a no-go argument for me. I am working on C-mannosylation and these regularly come in two or three on a peptide to which you can have a fucose in addition. Is this solved by now? I don't need to call it glyco, is this the case in any search? I usually add oxydation of M and W, deamination (N and Q), both stable, fucose on S or T (labile) and Hexose on W (with a partially labile fragment). This can be many modifications.

Thank you for any information, Hans

[dpolasky]

Hi Hans,
We have indeed now implemented O-Pair search in FragPipe, which supports searching for multiply O-glycosylated peptides if you have both collisional (HCD/CID) and electron-based (ETD/EThcD/etc) scans in your data. It is currently only searching S/T as the possible glycosites, but we can easily add an option to specify which amino acids to consider - I'll add that to the next release.

We do not currently have an option for HCD-only localization of multiple labile modifications on a peptide given the challenges in doing so, but you can still identify the peptide and total mass of modifications using a labile search.
Best,
Dan

[HansABakker]

Dear Dan, thank you for the quick response.
We are most interested in C-mannosylation of tryptophans. This is in principle stable (+162 Da), but can lose 120 Da by cross ring cleavage. In MS2 spectra you usually see both (the +162 Da usually still stronger), which is extremely good evidence for this modification. We can have 3 C-mannosylations in a peptide like xxWxxWxxWxK. In y-ion with two mannoses, you then have +324 Da and two extra masses with -120 and -240. I had proteome discoverer 3.0 as tryout version for the last months that was able to show me this nicely in resulting MS spectra. I did not manage to do this in our 2.4 version. I am not able to add neutral losses of modifications there, although I think this should be a rather standard feature. Would MS-Fragger be able to do this and mark these in the spectra? Or even give the spectra points for this. I now get many spectra with high quality values, but if I look at it, I see that there is somehow 162 Da added to the peptide (probably a labile O-mannose), but it is certainly not sitting on the tryptophan.
I usually do not add fucose in the search, unless I look for one specific peptide only. It becomes a mess: Fucose is a desoxyhexose and combining oxidation, hexose and desoxyhexose in a search will give you a lot of wrongly annotated spectra. You then have to leave out oxidation.
We occasionally use ETD, but prefer HCD for C-mannosylation.
Best regards, Hans

[dpolasky]

Hi Hans,
I see - this actually sounds like a great case for our new labile search mode (https://www.sciencedirect.com/science/article/pii/S1535947623000488), which can use what we call "fragment remainder ions" (the 120 loss in this case) to localize the modification in HCD spectra. It is based on a mass offset search, which means it is not designed for looking at multiple labile modifications on one peptide, but there is a workaround by specifying (for example) two mannosylations as variable modifications and one as a labile mass offset. This is similar to what we did for the phospho searches in the paper. That would be my recommendation for now, though hopefully we will be able to support multiple-labile localization in the future.

Best,
Dan

[HansABakker]

Dear Dan,
Thanks, that sounds like what I need, I will have a look at the paper and downlood the software. I see the paper is rather new. Indeed, I already looked for methods to spot the 120 Da losses. I will let you know how this works out.
Best, Hans

[fcyu]

The current version of FragPipe has O-pair localization module that supports multiple O-glycans. I think this issue can be considered as resolved.

Best,

Fengchao