Has anyone successfully used CAMERA for peptide MS data?
Peptides have different isotopic distributions from metabolizes, due mostly (I believe) to their generally larger masses and higher charge states. I tried CAMERA with default settings on xcms centWave-picked peaks, hoping to get isotopic clusters annotated. But the annotations were mostly 1+, with a smaller number of 2+ Isotopes (perhaps a few 3+). Whereas my peptides run up to 5+, with the majority of them 2+ and 3+.
First, can this be done? And if so, is it done by modifying the isotopeMatrix? From earlier postings, it appears this only handles one charge state. Do I construct multiple matrices, one per charge state?
Thank you,
- Peter
P.S. This is related to my recent question on the xcms forum, where Jan suggested using CAMERA.
Hi Peter,
CAMERA has not been written with peptides in mind, so at least the defaults need to be modified:
1) maxcharge=3, maxiso=4 in findIsotopes() need to be increased. Maybe it would help to
disable filter: Should C12/C13 filter be applied which makes an intensity check between M and M+1.
also, make sure the mass accuracy is stringent enough, the default mzabs=0.01 is too relaxed for e.g. an Orbitrap.
Having said that, I'd expect an isotope detection with a peptide intensity model
to be better suited for proteomics data. If you still want to use (and possibly improve)
CAMERA, I'd be happy to help, time permitting.
Yours,
Steffen
Hi, Steffen,
Thanks for your reply. Is the intensity model something I can fiddle with? Or is that deep in the code? In my particular case, I'm dealing with glycopeptides, so I would probably need a somewhat different intensity model than straight peptides... a blend of peptide and glycan intensity models, with knowledge of the contributions of each (which in my case I can derive from my de novo glycan identification software, but which others may not have access to).
In the meantime I've started in on my own isotope grouping approach, which is showing some promise. Unless I can tweak CAMERA's intensity model myself (which also needs to be mass-dependent), I will probably not be able to use the tool.
Thanks,
- Peter
Hi,
findIsotopes() has
isotopeMatrix: four column m/z-diff and ratio Matrix, for matching
isotopic peaks.
Yours,
Steffen