By integrating the simplicity of 18O-labeling and the low signal-to-noise of MS/MS spectra with supporting software and combining them with global shotgun protein identification, a robust quantitative pipeline has been created that avoids the disadvantages of other quantitative approaches. Test mixtures of labeled and unlabeled peptides were subjected to LC-MS/MS profiling experiments. Software programs were developed and applied to automatically determine protein ratios between two samples while applying a correction for incomplete labeling. The measurement of relative abundance at the product ion (MS/MS) level, instead of at the full scan (MS) level, is shown to provide excellent accuracy and sensitivity. Ratio distributions approached the expected means, allowing empirical derivation of confidence level cutoffs for determining statistically significant fold-changes in protein abundance. A set of stringent criteria for detecting spurious ratios based on consistency checking between unlabeled and labeled y-ion pairs was found to highlight putative false positive identifications.
| Identifer | oai:union.ndltd.org:TORONTO/oai:tspace.library.utoronto.ca:1807/25509 |
| Date | 30 December 2010 |
| Creators | White, Carl |
| Contributors | Emili, Andrew |
| Source Sets | University of Toronto |
| Language | en_ca |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.0018 seconds