Computational methods and tools for protein phosphorylation analysis

Abstract

Signaling pathways represent a central regulatory mechanism of biological systems where a key event in their correct functioning is the reversible phosphorylation of proteins. Protein phosphorylation affects at least one-third of all proteins and is the most widely studied posttranslational modification. Phosphorylation analysis is still perceived, in general, as difficult or cumbersome and not readily attempted by many, despite the high value of such information. Specifically, determining the exact location of a phosphorylation site is currently considered a major hurdle, thus reliable approaches are necessary for the detection and localization of protein phosphorylation. The goal of this PhD thesis was to develop computation methods and tools for mass spectrometry-based protein phosphorylation analysis, particularly validation of phosphorylation sites. In the first two studies, we developed methods for improved identification of phosphorylation sites in MALDI-MS. In the first study it was achieved through the automatic combination of spectra from multiple matrices, while in the second study, an optimized protocol for sample loading and washing conditions was suggested. In the third study, we proposed and evaluated the hypothesis that in ESI-MS, tandem CID and HCD spectra of phosphopeptides can be accurately predicted and used in spectral library searching. This novel strategy for phosphosite validation and identification offered accuracy that outperformed the other currently existing popular methods and proved applicable to complex biological samples. And finally, we significantly improved the performance of our command-line prototype tool, added graphical user interface, and options for customizable simulation parameters and filtering of selected spectra, peptides or proteins. The new software, SimPhospho, is open-source and can be easily integrated in a phosphoproteomics data analysis workflow. Together, these bioinformatics methods and tools enable confident phosphosite assignment and improve reliable phosphoproteome identification and reporting