Theoretical peptide z-ion mass appears to be incorrect

support
Theoretical peptide z-ion mass appears to be incorrect JHamey  2018-11-14
 

Hi there Skyline team,

I've come across a problem to do with the theoretical mass of z-ions in Skyline. They appear to be consistently off by 1 Da - they are predicted to be 1 Da lighter than they actually are. E.g. a z13 ion is predicted to be 1316.7 Da when it is actually 1317.7 Da. When I searched for threads related to this issue I came across this one (Issue 498):

https://skyline.ms/issues/home/issues/details.view?issueId=498&_docid=issue%3A498

It appears as if the "fix" that was made to make z-ions 1 Da lighter was done in error. The 1 Da heavier masses for z-ions I am getting from MS-Product match what I see experimentally for a normal peptide (it was noted in that thread that the peptide where the "error" was observed contained an abnormal gamma linkage at the position where the mass shift was thought to be wrong). C-ion masses are perfectly ok. I see this consistently for at least 5 different z-ions in the peptide.

Thank you in advance for your help.

 
 
Nick Shulman responded:  2018-11-14
The numbers calculated in Skyline match what's on this web site:
http://db.systemsbiology.net:8080/proteomicsToolkit/FragIonServlet.html
They also match what's in the supplemental data for this paper:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4095805/

Do you have another tool which calculates different numbers for Z-ions? Can you send us a screenshot of that?

Skyline subtracts an NH3 from a y-ion in order to get the mass of the z-ion. When I look at the molecule diagram on the Wikipedia page, I can't figure out why that would be the correct thing to do, but it does match all of the results that I can find online.

If you'd like, you can send us your Skyline document. There might be something else going on.
In Skyline, you can use the menu item:
File > Share > (complete)
to create a .zip file containing your Skyline document and supporting files, including extracted chromatograms.

If that .zip file is less than 50MB then you can attach it to this support request. Otherwise, you can upload that here:
https://skyline.ms/files.url

You should also send us your raw file so we can see those MS2 spectra.
-- Nick
 
JHamey responded:  2018-11-14
Thanks for the quick response Nick.

I got the masses for z-ions from MS-Product (http://prospector.ucsf.edu/prospector/cgi-bin/msform.cgi?form=msproduct). See the attached image for the spectrum with matching z-ions.

However, I just went and searched my data with Mascot and it appears what MS-Product calls z-ions are in fact z+1 ions (see attached image). Matrix science even says that ETD generally produces z+1 ions (http://www.matrixscience.com/help/fragmentation_help.html) and so Mascot didn't even search for z-ions. So it appears as if the mass calculation for z-ions is correct, but the problem is that what ETD produces (at least for lower charge state peptides) is mostly z+1 ions. So is it possible to use z+1 ions instead of z-ions?

Thank you.
 
Brendan MacLean responded:  2018-11-18
The Mascot image you sent appears to have z+2 also. It looks like MS-Product allows c+2, c+1, c, c-1 and z, z+1, z+2, z+3. So, maybe the right solution is to also add these to the list of ions that Skyline supports. If so, probably we should also add a pop-up checklist attached to a menu button like we have for chemical formulae. Do you have any more information on the need all 8 c- and z-ion options? Or should we contact the authors of Mascot and MS-Product for a better sense of the complete range we should support for c- and z-ions?

Thanks for pointing this out to us.

--Brendan
 
Brendan MacLean responded:  2018-11-18
I read the page from Matrix Science where they say, "Electron Transfer Dissociation and Electron Capture Dissociation generate primarily c, y, z+1, and z+2 ions. In some cases, w ions can also be prominent." So, maybe ETD users only need z+1 and z+2 added to Skyline? And we can wait for some other use case to require the other 4 c and z ions (c+2, c+2, c-1, z_3).

Does that seem like it would be enough for you to achieve what you were hoping for in your Skyline use for this experiment?

--Brendan
 
JHamey responded:  2018-11-18
That would be fantastic - certainly at least having z+1 and z+2 would be really good. Certainly I do see at least c+1 and c+2 ions, but since there's always the c ion as well it's not super important from what I've seen, whereas since it appears that the z+1 ion appears without the corresponding z ion, having that added would be really useful.

Thanks for the help with this.
 
Brendan MacLean responded:  2018-11-20
Okay. I think I understand a bit better after running this by David Creasy at Matrix Science (Mascot developer). There is a difference in nomenclature between Matrix Science (Biemann) and Protein Prospector (UCSF) which is explained by UCSF here:

http://prospector.ucsf.edu/prospector/html/misc/faq.htm#z_ions

With some further elaboration of how these ions show up in ETD experiments here:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2836188/#R10

Where the paper refers to Biemann z+1 and z+2 (and c and c-1) in their tables when they use z• and z+1. So, in Matrix Science/Biemann nomenclature (which Skyline seems to have adopted on March 1, 2017 after a prior support request to change z-ion calculation from -NH2 to -NH3), when you have z+2, you might also have c-1, but the probability of c-1 appears to be very close to noise (measured by x) in the tables of the paper.

I am not sure you are really seeing c+1 and c+2 as ions per se since the +1 and +2 are supposed to refer to hydrogen. The c+1 and c+2 you see, if you always see them with c, may simply be the heavy isotopes of the c-ions caused mostly by 13C in the molecule, and Skyline never targets those for any other ion type.

I think we will stay with Matrix Science/Biemann nomenclature to avoid flip-flopping on this, and we will try to soon add support for z+1 and z+2, which appear to be true monoisotopic ion forms that appear in ETD, though z+2, according to Devid Creasy, will only be seen with charge 3+ and above. Though, I am now curious what the nature of the experiment was that caused the prior support request when our z-ion calculation was actually calculating z+1, and Biemann z ions don't seem to appear in ETD and ECD.
 
JHamey responded:  2018-11-20
That all makes a lot of sense. I think you're completely right that the c+1, c+2 ions are just isotopic peaks. Their distribution looks like what I see in HCD/CID data too.

I think the previous request was unusual as it wasn't a normal peptide bond being analysed - so maybe this affected the likelihood of seeing z+1 ions instead of z• ions with ETD fragmentation at that bond.

Thanks again for looking into this.
 
bernhard blank responded:  2019-07-22
Hi everyone,

I'm piggybacking on this request rather than opening a new one, as I have exactly the same issue - I have peptides fragmented by EThcD and I'm primarily seeing z+1 ions (as by Mascot/Biemann nomenclature), but I cannot set the corresponding transitions in Skyline (Version 19.1).
Are the z+1 / z+2 ions already included and I just can't figure out how to set them or is this still work in progress?

Thanks a lot in advance and kind regards,
Bernhard
 
Brendan MacLean responded:  2019-07-22
Thanks for bringing this up again. Unfortunately, work on this issue has not yet started. I was thinking of just adding z+1 ions by allowing "z*" in the Transition Settings - Filter tab Ion Types field. You seem to be confirming this choice by saying "I'm primarily seeing z+1 ions". But, please let me know if you feel we also really need z+2 ions.

Skyline originally supported z+1 ions as "z" until a user contacted us and told us our z-ion calculation was off by 1, and we changed the calculation for Skyline 3.7:

https://skyline.ms/notes.url#skyline_3_7

You can get Skyline releases back to 1.4 from the "Unplugged" installers page, by clicking "I Agree" and then "Archive".

https://skyline.ms/skyline64zip.url

So, you can still get a 3.6 installer that should allow you to monitor z+1 ions (but then not true z ions), if Skyline 3.6 still has the other features you need.

I have seen a successful instance of using Skyline for EThcD, probably with a version before we switched to calculating true z-ions and lost the ability to calculate z+1 ions, and before we knew about this difference.

Hope this helps and we will try to make it possible soon to target both z- and z+1-ions in the same instance of Skyline. Please let us know if you feel we also need to support z+2 (or any type of c+1 or c+2 ions). Thanks for your feedback, and thanks for including it on this existing thread.

--Brendan
 
bernhard blank responded:  2019-07-23
Hi Brendan,

thanks a lot for your quick respond! That's indeed a good workaround to go for the archived 3.6 version when the z-ion calculation was done differnently. I see my z+1 ions simply as z-ions now, which is fine. Everyhting else seems to work as well with the older version, so I guess that problem is solved for me at the moment!

As for the general occurence of the individual ion types I can just make assumptions based on a limited number of observations. In my EThcD spectra at least the main c-terminal ions are z+1, with z+2 getting more intense with higher precursor charge states (3 and above) and m/z. So I guess it would make sense to include both types in a future release. As for n-terminal ions I only see c ions, c+1 or c+2 are usually not observable in my data.

Thanks again for the resposne and the great work you guys are doing! Highly appreciated!

Kind regards,
Bernhard