The recent availability of fully sequenced genomes from multiple related organisms has allowed researchers to investigate many long-standing questions in biology using the emerging paradigm of comparative genomics. These include: how phenotypic novelty arises through sequence divergence, how genome organization is restructured in drastic and subtle ways, and how the rich repertoire of interactions that constitute gene regulatory networks exhibit tremendous plasticity.

Comparative genomics relies heavily on determining the corresponding features between organisms. An accurate and comprehensive cross-species mapping is essential for extracting meaningful insights by comparing genome sequences. For example, tracing the exact evolutionary history of each gene in a panel of species allows us investigate the consequences of gene duplication and loss events; two key sources of phenotypic innovation.

In this thesis, we propose computational methods for advancing these biological questions. First, we introduce a new framework for identifying orthologous genes across a phylogeny, capturing exactly when genes were duplicated and lost. This general framework integrates various sources of evidence into a robust and scalable approach for recovering accurate gene histories from the genome annotations of extant species. We apply this framework across a panel of sequenced Ascomycota fungi genomes, allowing us to pursue new insights into the evolutionary signatures of gene duplication and loss in these species. We determine the functional principles that govern these processes, and how the forces of selection can operate distinctly between genes with different functional roles and between duplication events of different scales. Lastly, we extend this framework towards tracing the evolution of cis-regulatory elements. After building comprehensive species-specific motif catalogs, we map the corresponding elements between genomes, allowing us to glimpse into how transcription factor binding sites can be reshaped across evolution.

These studies thoroughly embrace the potential of comparative genomics. Sequencing and annotating genomes were the initial steps towards harnessing evolution to garner new discoveries about cellular behavior. Our precise mapping of shared ancestral relations between the components of each genome realizes this promise of comparative genomics. We study how the resulting catalog of gene phylogenies can reveal meaningful insights into the evolutionary constraints that govern them. In addition, by generating an encyclopedic library of functional cis-elements, we establish the foundations for future studies of regulatory network evolution.