search

smith-chem-wisc / FlashLFQ

Ultra-fast label-free quantification algorithm for mass-spectrometry proteomics
GNU Lesser General Public License v3.0
19 stars 14 forks source link

quantification of glycopeptides with multiple peaks #79

Closed Dmorgen closed 5 years ago

Dmorgen commented 5 years ago

Hi Guys,

I'm doing quite a bit of glycopeptidomics and we've noticed that there are multiple peaks per ID, most probably due to structural differences in the glycans. can you detect and quantify them separately?

Cheers, D.

trishorts commented 5 years ago

Rob will comment on this I'm sure. We have some nice updates on the way for metamorpheus that will improve coverage and quantification of glycopeptides. Some of these updates are in a pull request. So, hopefully within a month you'll see something. I'll try to remember to send you an email so you'll have a chance to see it in action.

Dmorgen commented 5 years ago

Great, thanks!

lonelu commented 5 years ago

Hi D, Do you mean the same glycopeptide has more than one MS1 features? Can you show me an example? Thanks, Lei

Dmorgen commented 5 years ago

Here one example - all three peaks have MS2 spectra that correspond to the sample glycopeptide. image

rmillikin commented 5 years ago

Some peaks are combined together into one peak in FlashLFQ to avoid noisy, near-baseline peaks from being double-counted, which throws off the intensity measurement for a particular peptide, since FlashLFQ uses peak height instead of area for quantification. The tolerance for combining peaks is 5 min, so every peak with the same modified sequence within 5 min of each other is combined into one. I can make this into a user-editable setting. You can also make a workaround with the current code by making the full sequence of each ID different. For example you could do:

Base Sequence: REEQFNTTFR
Full Sequence REEQFN(Glyco)TTFR_1

Base Sequence: REEQFNTTFR
Full Sequence REEQFN(Glyco)TTFR_2

Then each of those would be reported separately in the Peaks file (and in the Quantified Peptides file). I know it's a little annoying, but it'll work.

Dmorgen commented 5 years ago

thanks!

rmillikin commented 5 years ago

FYI, I've changed this in the upcoming FlashLFQ version. I've decided to not merge peaks together unless they have the same apex. Should help make this kind of thing easier to interpret, but it will probably report several peaks for peptides that bounce in and out of the limit of detection.