Modelling protein network effects of mutations and post-translational modifications in yeast

The aim of this thesis is to contribute to better understanding of essential molecular mechanisms in yeast Saccharomyces cerevisiae cells. Two main research questions are addressed: i. how do post-translational modifications (PTMs) affect yeast protein-protein interactions, and ii. what are the mechanisms underlying yeast ethanol tolerance. Post-translational modifications, such as the addition of chemical moieties on specific amino acids, provide an elegant way of regulating protein roles by affecting their localization, stability, and interactions. The last decades have witnessed major advances in mass spectrometry technologies, which have resulted in proteome-wide detection of PTM sites for a number of organisms. While this abundance of data is publicly available in numerous dedicated databases, the question remains as to what are the functional roles of these modifications. The existing computational solutions have tackled this problem by using a set of assumptions, for instance that PTMs introduce no or only local conformational changes. In addition, they face a number of limitations when it comes to PTM types, crosstalk, and position with respect to protein-protein interface. In this dissertation, the problem of predicting PTM roles in protein-protein interactions is approached in a more biologically realistic way. To this end, a prediction workflow centered around molecular dynamics simulations and free energy calculations (termed MD-MM/GBSA) is developed and benchmarked against the current methods. Alongside its better performance, MD-MM/GBSA comes with virtually no restrictions of PTM types, number, and location within the complex, all at a cost of computational time. The MD-MM/GBSA workflow is then applied to predict effects of lysine acetylation and serine/threonine/tyrosine phosphorylation on interactions of subunits in: i. three large protein complexes conserved among eukaryotes, namely the exosome, RNA polymerase II, and proteasome, and ii. all yeast protein complexes with high-quality protein structures available. One of the main observations is that the introduction of acetylation on the interaction interface typically has locally stabilizing contribution to binding, while the opposite is true for phosphorylation. However, while the prediction of local effects appears to be straightforward, co-occurrence of PTMs and their long-range conformational effects render the prediction of overall effect on protein-protein binding more complex. The s