When validating glycans, it is often useful to first take a step back and look at the results in context. At the highest level, make sure you are using the best database and have optimized your search settings. Once you are happy with your search and start reviewing your results, analyzing those results in a larger context is an important first step.
A good way to get started is to sort your results first by peptide, and second by retention time. Don’t forget that you can group the charge states of peptides together to see if the distribution is as expected. Then you can take a first pass at your data as an ordered peptide group. Glycans are constructed following synthetic pathways, so if you check the group, do they seem they related to a rational construction pathway? Also, ensure that retention times make sense for the glycan. Finally, inspect the MS2s of ambiguous assignments to ensure the correct glycan was assigned. Glycosylated peptides often yield poor fragmentation of the peptide backbone, so relying solely on peptide scores for validation can result in overlooking true positive spectra. For that reason, it is often necessary to include manual validation of MS2 spectra as part of the assessment of low scoring glycopeptide results. Remember that in the search settings, you can specifically select a ‘boost’ by using the ‘Show all N-glycans’ checkbox.
Setting up the search:
- Check out the Help→Open User Guides link (which opens our ‘resources’ web page) in the software to access two great application notes on setting up N and O-Glycan searches (or contact support@proteinmetrics.com).
- Select the correct database: one tissue expression system may produce different glycans than another system. Byonic includes several different glycan databases that you can choose from.
- This is especially important because one of the most common causes for error is that the right answer was not part of the search due to a missing glycan in the selected glycan database.
- Remember that our databases are open and easily edited and extended to meet your organization’s needs.
Examine Peptide Groups:
- Sort your results by peptide and retention time.
- Look at the top scoring result in a group and make sure all the other results make sense in that context. Do the glycans identified in a peptide group build upon one another?
- Keep in mind that you will often see high mannose and complex sialated forms together at the same site.
- With data arrayed by elution time, throw out results where the retention time is wrong.
- NeuAc and other sialic acids can result in a couple of minutes shift to more acidic areas of the chromatogram. Do you see that happening in the data? If not, there may be an issue with the identification.
- Evaluate other elution times clues such as the presences of doublet peaks when a substituent like a sialic acid moves from one antenna to another.
Evaluating the MS2:
- In most collisional glycopeptide MS2 spectra you should see signals for HexNAc (204) and HexNAcHex (366). If you do not see the fragments, but have an identification that includes these structural units, you may not be looking at a glycopeptide.
- Be aware of isobaric species that may lead to confusing identifications
- For smaller glycans, while mass does imply composition, there may be multiple structures for a given composition.
- As glycans increase in size and the mass is greater than about 4000 Th, there may also be different compositions with close masses (< 0.1 Th apart) that may complicate the identification.
- NeuAc can be confused with two fucose units, or NeuGc may be assigned rather than NeuAc on an oxidized peptide, so you should check for the presence of these fragments (274 and 290) if they are expected. Note, newer versions of the software (3.7 and greater) add a penalty for NeuAc-containing assignment without 274, and for NeuGc-containing assignment without 290 to minimize the occurrence of this.
- Take advantage of the Delta Scores and Delta Mod Scores as well as our new feature which allows visibility of alternative identifications if available.
- Carefully examine spectra with low Delta Scores as they indicate that Byonic found a different solution that explains the spectrum nearly as well.
- Assure that the N and C termini of the peptide are correct. If they are not correct, a purported truncated peptide with a glycan may be a full-length peptide with a different glycan.
- Look for b- and y- ions that cover the termini.
- Alternatively, look for the intact peptide with a small glycan fragment still attached, this is also a clue that we have the correct peptide as well as a help in localization. With HCD / QTOF fragmentation at the correct energy, bare peptide (y0) and bare peptide + base GlcNAc or GalNAc (y1) are reliable peaks.
- Look for fragment ions that contain the glycan or part of the glycan. These fragments can help confirm the localization of the glycosylation.
- Be mindful of fragments labeled with a tilde (eg ~y5) which represent fragments that have lost the glycan (or other labile species) entirely. While this is reasonable and expected, in the absence of other confirmatory information, it may also indicate a site localization issue.
A few sample MS2 spectra:
