A high dimensional map of phosphorylation-dependent signaling in budding yeast

Abstract

Development of mass spectrometry-based phosphoproteomics has substantially expanded our understanding of cellular processes regulated by protein phosphorylation. However, most phosphoproteomic studies are low dimensional (<5 conditions) due to analytical limitations. Phosphoproteomic profiling at much higher dimensionality is required to obtain systematic insight into the architecture and function of phosphorylation-dependent signaling networks. Here we present a high-dimensional quantitative phosphoproteomic perturbation atlas in Saccharomyces cerevisiae at large scale (>100 perturbations, ~600 samples). We applied a combination of novel methods for automated sample-preparation, prioritization of functional phosphosites by machine learning and advanced data-analysis to overcome current limitations in large-scale phosphoproteomics. Integrating this massive quantitative resource gave us the unique opportunity to systematically elucidate signaling modules, profile kinase activities, and predict functional phosphorylation sites at the organismal scale.