Over the past two decades, high-throughput (HTP) technologies such as microarrays and mass spectrometry have fundamentally changed the landscape of aging and disease biology. They have revealed novel molecular markers of aging, disease state, and drug response. Some have been translated into the clinic as tools for early disease diagnosis, prognosis, and individualized treatment and response monitoring. Despite these successes, many challenges remain: HTP platforms are often noisy and suffer from false positives and false negatives; optimal analysis and successful validation require complex workflows; and the underlying biology of aging and disease is heterogeneous and complex. Methods from integrative computational biology can help diminish these challenges by creating new analytical methods and software tools that leverage the large and diverse quantity of publicly available HTP data. In this thesis I report on four projects that develop and apply strategies from integrative computational biology to identify improved biomarkers and therapeutics for aging and disease. In Chapter 2, I proposed a new network analysis method to identify gene expression biomarkers of aging, and applied it to study the pathway-level effects of aging and infer the functions of poorly-characterized longevity genes. In Chapter 4, I adapted gene-level HTP chemogenomic data to study drug response at the systems level; I connected drugs to pathways, phenotypes and networks, and built the NetwoRx web portal to make these data publicly available. And in Chapters 3 and 5, I developed a novel meta-analysis pipeline to identify new drugs that mimic the beneficial gene expression changes seen with calorie restriction (Chapter 3), or that reverse the pathological gene changes associated with lung cancer (Chapter 5). The projects described in this thesis will help provide a systems-level understanding of the causes and consequences of aging and disease, as well as new tools for diagnosis (biomarkers) and treatment (therapeutics).

Eukaryotic DNA-binding proteins mediate many important steps in embryonic development and gene regulation. Consequently, a better understanding of these proteins would hopefully allow a more complete picture of gene regulation to be determined. In this study, a structure-based subfamily classification of the homeodomain family of DNA-binding proteins was undertaken in order to determine whether sub-groupings of a protein family could be identified that corresponded to differences in specific function, and identification of subfamily-determining residues was performed in order to gain some insight on functional differences via analysis of the residue properties. Subfamilies appear to have different specific DNA binding properties, according to DNA profiles obtained from TRANSFAC [1] and other sources in the literature. Subfamily-specific residues appear to be frequently associated with the protein-DNA interface and may influence DNA binding via interactions with the DNA phosphate backbone; these residues form a conserved profile uniquely identifying each subfamily.

Eukaryotic DNA-binding proteins mediate many important steps in embryonic development and gene regulation. Consequently, a better understanding of these proteins would hopefully allow a more complete picture of gene regulation to be determined. In this study, a structure-based subfamily classification of the homeodomain family of DNA-binding proteins was undertaken in order to determine whether sub-groupings of a protein family could be identified that corresponded to differences in specific function, and identification of subfamily-determining residues was performed in order to gain some insight on functional differences via analysis of the residue properties. Subfamilies appear to have different specific DNA binding properties, according to DNA profiles obtained from TRANSFAC [1] and other sources in the literature. Subfamily-specific residues appear to be frequently associated with the protein-DNA interface and may influence DNA binding via interactions with the DNA phosphate backbone; these residues form a conserved profile uniquely identifying each subfamily.

Over the past two decades, high-throughput (HTP) technologies such as microarrays and mass spectrometry have fundamentally changed the landscape of aging and disease biology. They have revealed novel molecular markers of aging, disease state, and drug response. Some have been translated into the clinic as tools for early disease diagnosis, prognosis, and individualized treatment and response monitoring. Despite these successes, many challenges remain: HTP platforms are often noisy and suffer from false positives and false negatives; optimal analysis and successful validation require complex workflows; and the underlying biology of aging and disease is heterogeneous and complex. Methods from integrative computational biology can help diminish these challenges by creating new analytical methods and software tools that leverage the large and diverse quantity of publicly available HTP data. In this thesis I report on four projects that develop and apply strategies from integrative computational biology to identify improved biomarkers and therapeutics for aging and disease. In Chapter 2, I proposed a new network analysis method to identify gene expression biomarkers of aging, and applied it to study the pathway-level effects of aging and infer the functions of poorly-characterized longevity genes. In Chapter 4, I adapted gene-level HTP chemogenomic data to study drug response at the systems level; I connected drugs to pathways, phenotypes and networks, and built the NetwoRx web portal to make these data publicly available. And in Chapters 3 and 5, I developed a novel meta-analysis pipeline to identify new drugs that mimic the beneficial gene expression changes seen with calorie restriction (Chapter 3), or that reverse the pathological gene changes associated with lung cancer (Chapter 5). The projects described in this thesis will help provide a systems-level understanding of the causes and consequences of aging and disease, as well as new tools for diagnosis (biomarkers) and treatment (therapeutics).

Protein phosphorylation is a prevalent reversible post-translational modification that influences protein functions. The advent of phospho-proteomic technologies now enables proteome-wide quantitative detection of residues phosphorylated under different physiological conditions. The functional consequences of the majority of these phosphorylation events are unknown. This calls for endeavors to characterize their molecular functions and cellular effects. This can be facilitated by systematic approaches to categorize phosphorylation events, interpret their importance and infer their functions. I carried out comparative, evolutionary and integrative analyses on in vivo phosphorylation events to address these challenges. First, I performed cross-species comparative phospho-proteomic analysis to identify evolutionarily conserved phosphorylation events in human. A sequence alignment approach was used to identify phosphorylation events conserved at similar sequence positions across orthologous proteins and a network alignment approach was applied to identify potential evolutionarily conserved kinase-substrate interactions. Conserved human phosphoproteins identified are found enriched for proteins encoded by known cancer- and disease-associated genes. Next, I developed a new approach to analyze the sequence conservation of known phosphorylated residues on human, mouse and yeast proteins that factored in the background mutational rates of protein and phosphorylatable residue. Furthermore, sites were analyzed according to (i) characterized functions, (ii) prevalence, (iii) stoichiometry, their occurrence in (iv) structurally disordered/ordered protein regions, in (v) proteins of various abundance and in (vi) proteins with different protein interaction propensity to identify the factors influencing sequence conservation of phosphorylated residues. Importantly, my analysis suggests that false positives and randomly phosphorylated residues are present in existing phosphorylation datasets and they are more common on high abundance proteins. Lastly, I characterized the theoretical maximum phosphorylation capacity in terms of phosphorylatable residues and discovered that genomic tyrosine frequency correlates negatively and significantly with tyrosine kinase frequency and cell type in metazoan. This observation suggests that fidelity of phosphotyrosine signaling occurred partially through global tyrosine depletion.

Global quantitative analysis of genetic interactions provides a powerful approach for deciphering the roles of genes and mapping functional relationships amongst path-ways. Using colony size as a proxy for fitness, I developed a method for measuring ge-netic interactions from high-density arrays of yeast double mutants generated by synthet-ic genetic array (SGA) technology. I identified several experimental sources of systematic variation and developed normalization strategies to obtain accurate fitness measurements. I used this scoring method to map quantitative genetic interactions among 5.4 million yeast double mutants and generated the first functionally unbiased genetic interaction map of a eukaryotic cell. My map produced an unprecedented view of the cell in which genes of similar biological processes cluster together in coherent subsets and functionally interconnected bioprocesses map next to each other. We discovered several physiological and evolutionary gene features that are characteristic of genetic interaction hubs, and explored the relationship between genetic and protein-protein interaction networks. In particular, by comparing quantitative single and double mutant phenotypes, we identified specific cases of positive genetic interactions, termed genetic suppression, and constructed a global network of suppression interactions among protein complexes. I also demonstrated that an extensive and unbiased mapping of genetic interactions provides a key for interpreting chemical-genetic interactions and identifying drug targets. In addition, I used genome-wide SGA data to map profiles of genetic linkage along all sixteen yeast chromosomes. These linkage profiles recapitulated previously identified recombination patterns and uncovered an unexpected correlation between chromosome length and the extent of centromere-related recombination repression. These findings suggest a chromosome size-dependent mechanism for ensuring proper chromosome segregation and highlight the SGA methodology as a unique approach for systematic analysis of yeast meiotic recombination.

Pulmonary metastasis is the most frequent cause of osteosarcoma (OS) mortality. The aim of this study was to discover and characterize genetic networks differentially expressed in metastatic OS. Supervised network analysis of OS expression profiles was performed to discover genetic networks differentially activated or organized in metastatic OS. Broad trends among the profiles of metastatic tumours included aberrant activity of intracellular organization and translation networks, as well as disorganization of metabolic networks. The differentially activated PRKCε-RASGRP3-GNB2 network, which interacts with the disorganized DLG2 hub, was additionally found to be differentially expressed among in vitro models of human OS metastasis. PRKCε transcript was more abundant in some metastatic OS tumours; however the difference was not significant overall. In functional studies, PRKCε was not found to be involved in migration of M132 OS cells, but its protein expression was induced in M112 OS cells following IGF-1 stimulation.

Pulmonary metastasis is the most frequent cause of osteosarcoma (OS) mortality. The aim of this study was to discover and characterize genetic networks differentially expressed in metastatic OS. Supervised network analysis of OS expression profiles was performed to discover genetic networks differentially activated or organized in metastatic OS. Broad trends among the profiles of metastatic tumours included aberrant activity of intracellular organization and translation networks, as well as disorganization of metabolic networks. The differentially activated PRKCε-RASGRP3-GNB2 network, which interacts with the disorganized DLG2 hub, was additionally found to be differentially expressed among in vitro models of human OS metastasis. PRKCε transcript was more abundant in some metastatic OS tumours; however the difference was not significant overall. In functional studies, PRKCε was not found to be involved in migration of M132 OS cells, but its protein expression was induced in M112 OS cells following IGF-1 stimulation.

The shikimate pathway links carbohydrate metabolism to the biosynthesis of the aromatic amino acids and an enormous variety of aromatic compounds with essential functions in all kingdoms of life. Aromatic compounds derived from the plant shikimate pathway have substantial biotechnological value and many are essential to the diet of metazoans whose genomes do not encode shikimate pathway enzymes. Despite its importance to the physiology of plants and human health the regulatory mechanisms of the plant shikimate pathway are not well understood. Shikimate kinase (SK) genes encode an intermediate step in the shikimate pathway and were previously implicated in regulation of the plant shikimate pathway. The distribution of SK genes in higher plants was resolved using phylogenetic and biochemical methods. The two SK isoforms of Arabidopsis thaliana, AtSK1 and AtSK2, were functionally characterized. AtSK1 expression is induced by heat stress and the recombinant enzyme was shown to form a homodimer which is important for maintaining the stability and activity of the enzyme at elevated temperatures. The crystal structure of AtSK2, the first reported plant SK structure, identified structural features unique to plant SKs which may perform important regulatory functions. The resolution of bona fide SKs in higher plants led to the discovery of two novel neofunctionalized homologs - Shikimate Kinase-Like 1 (SKL1) and SKL2. These novel genes evolved from SK gene duplicates over 400 million years ago and are found in all major extant angiosperm lineages, suggesting they were important in the development of biological properties required by land plants. The description of albino and variegated skl1 mutants in Arabidopsis thaliana implicate the SKL1 gene product as an important regulator of chloroplast biogenesis. Functional assays were attempted to determine the biochemical function of SKL1 and recombinant constructs of the Arabidopsis thaliana SKL1 protein were crystallized towards structure determination. The results of this thesis further our understanding of the organization and regulation of the plant shikimate pathway. Furthermore, the discovery of SKL1 may yield important insights into chloroplast biogenesis and function. The evolution of the plant SK superfamily highlights the utility of SKs as scaffolds for functional innovation.

The currently known network of human protein-protein interactions (PPIs) is providing new insights into diseases and helping to identify potential therapies. However, according to several estimates, the known interaction network may represent only 10% of the entire interactome - indicating that more comprehensive knowledge of the interactome could have a major impact on understanding and treating diseases. The primary aim of this thesis was to develop computational methods to provide increased coverage of the interactome. A secondary aim was to gain a better understanding of the link between networks and phenotype, by analyzing essential mouse genes. Two algorithms were developed to predict PPIs and provide increased coverage of the interactome: FpClass and mixed co-expression. FpClass differs from previous PPI prediction methods in two key ways: it integrates both positive and negative evidence for protein interactions, and it identifies synergies between predictive features. Through these approaches FpClass provides interaction networks with significantly improved reliability and interactome coverage. Compared to previous predicted human PPI networks, FpClass provides a network with over 10 times more interactions, about 2 times more proteins and a lower false discovery rate. This network includes 595 disease related proteins from OMIM and Cancer Gene Census which have no previously known interactions. The second method, mixed co-expression, aims to predict transient PPIs, which have proven difficult to detect by computational and experimental methods. Mixed co-expression makes predictions using gene co-expression and performs significantly better (p < 0.05) than the previous method for predicting PPIs from co-expression. It is especially effective for identifying interactions of transferases and signal transduction proteins. For the second aim of the thesis, we investigated the relationship between gene essentiality and diverse gene/protein features based on gene expression, PPI and gene co-expression networks, gene/protein sequence, Gene Ontology, and orthology. We identified non-redundant features closely associated with essentiality, including centrality in PPI and gene co-expression networks. We found that no single predictive feature was effective for all essential genes; most features, including centrality, were less effective for genes associated with postnatal lethality and infertility. These results suggest that understanding phenotype will require integrating measures of network topology with information about the biology of the network’s nodes and edges.