Properties of isotope patterns and their utility for peptide identification in large-scale proteomic experiments
Article first published online: 14 APR 2013
Copyright © 2013 John Wiley & Sons, Ltd.
Rapid Communications in Mass Spectrometry
Volume 27, Issue 9, pages 1067–1075, 15 May 2013
How to Cite
Okawa, S., Fischer, B. and Krijgsveld, J. (2013), Properties of isotope patterns and their utility for peptide identification in large-scale proteomic experiments. Rapid Commun. Mass Spectrom., 27: 1067–1075. doi: 10.1002/rcm.6551
- Issue published online: 10 APR 2013
- Article first published online: 14 APR 2013
- Manuscript Accepted: 17 FEB 2013
- Manuscript Revised: 14 FEB 2013
- Manuscript Received: 2 DEC 2012
In proteomic experiments, isotope patterns are routinely generated for all detected peptides. While this pattern is determined by peptide composition, it has not been evaluated as a parameter that can help in the process of peptide identification.
First, we investigated how the relative isotope abundance (RIA) accuracy in proteomic data sets depends on the spectral intensity, resolution, and the number of mass spectrometry (MS) 1 scans, using an Orbitrap Velos mass spectrometer. Next, we explored the discriminatory power of isotope patterns in the context of proteome analyses of various complexities, either alone or in combination with a Mascot database search. Finally, we provide a theoretical framework for the required accuracies of both peptide mass and RIA for peptide identification.
We demonstrate that the RIA error obtained in routine proteome analyses is 4–5%, and that this is only modestly influenced by spectral intensity, resolution, and the number of MS1 scans. While RIA alone has no discriminatory power, in a Mascot search isotope patterns can distinguish top scoring hits from runner-up hits in 70–95% of cases. Our theoretical approach shows that RIA accuracy needs to be ~0.2% in order to uniquely identify peptides in full proteomes.
Our results demonstrate that isotope patterns can have discriminatory power when used in combination with a classical database search. Inclusion of this parameter in proteomic workflows may help to increase confidence in peptide identification, but in practical terms this will be limited to small proteomes. Copyright © 2013 John Wiley & Sons, Ltd.