A comparative genomics approach to understanding sensory adaptation in mammalian evolution

Sensory perception is the ability to detect, interpret and respond to environmental cues. With roles in sourcing food, evading predators, and locating mates, sensory systems are likely to be at the forefront of evolution in mammals, allowing them to occupy and thrive in a variety of diverse and ecologically demanding environments. At the molecular level, mammalian sensory receptor genes exhibit remarkable copy number variation. This genetic variation is shaped by ecological niche, providing insight into the highly specialised and diverse sensory capabilities across Mammalia. While there has been an exponential increase in the number of available mammalian genome assemblies in recent years, current annotation pipelines are not optimised to capture the extensive variation within sensory receptor gene families, often underestimating the true number of genes in a given species. Moreover, naming and classification of these genes has remained a significant challenge in the field, owing to their rapid duplication rates and lineage specific nature. To overcome these limitations, new sensory receptor gene mining and annotation pipelines were developed and benchmarked in this thesis. By applying these tools to a taxonomically diverse dataset comprising over 550 mammalian species, subfamily-level classifications were achieved for over 700,000 receptors, enabling research into sensory receptor gene evolution at resolutions previously unattainable. Using this comprehensive dataset, the key factors shaping sensory adaptation and trade-offs were explored. Specifically, dietary niche, habitat, reproductive traits and predation were identified as drivers of sensory adaptation in mammals. Trade-offs were identified between visual and chemosensory systems in primates, and between echolocation and olfaction in bats. Across Mammalia, complex vocal communication emerged as a potential factor linked to a reduced reliance on olfactory perception, providing novel insights into the evolutionary history underlying speech and language acquisition in humans. While the dataset generated in this thesis serves as an essential starting point towards understanding the diverse sensory abilities of mammals, there is currently a lack of information regarding the ligand specificity of individual sensory receptors. Future research aimed at deorphaning these receptors may provide a better understanding of their functional roles, enabling a deeper exploration into the selective pressures driving sensory evolution.