Network analyses of proteome evolution and diversity
MetadataShow full item record
The mapping of biomolecular interactions reveals that the function of most biological components depends on a web of interrelations with other cellular components, stressing the need for a systems-level view of biological functions. In this work, I explore ways in which the integration of network and genomic information from different organizational levels can lead to a better understanding of cellular systems and components. First, studying yeast, I show that the evolutionary properties of target genes constitute the dominant determinant of transcription factor (TF) evolutionary rate and that this evolutionary modularity is limited to activating regulatory relationships. I also show that targets of fast-evolving TFs show greater evolutionary expression changes and are enriched for niche-specific functions and other TFs. This work highlights the importance of trans-regulatory network evolution in species-specific gene expression and network adaptation. Next, I show that genes either lost or gained across fungal evolution are enriched in TFs and have very different network and genomic properties than universally conserved genes, including, in sharp contrast to other networks, a greater number of transcriptional regulators. Placing genes in the context of their evolutionary life-cycle reveals principles of network integration of gained genes and evidence for the progressive network and functional marginalization of genes as an evolutionary process preceding gene loss. In the final chapter, I study how alternative splicing (AS)-driven expansion of human proteome diversity leads to system-level complexity through the AS-mediated rewiring of the protein-protein interaction network. By overlaying different network and genomic datasets onto the first large-scale isoform-resolution interactome, I found that differentiating between splice variants is essential to capturing the full extent of the network's functional modularity. I also discovered that AS-mediated rewiring preferentially affects tissue-specific genes and that topologically different patterns of rewiring have distinct functional consequences. Furthermore, I found that most rewiring can be traced to the AS of evolutionarily conserved sequence modules, which promote or block interactions and tend to overlap linear motifs and disrupt known domain-domain interactions. Together, this work demonstrates that a network-level perspective and genomic data integration are essential to understanding the evolution and functional diversity of proteomes.