Comparative analyses of adaptive phenomena in eukaryotic systems

All phenotypic adaptations are encoded in the genome, although untangling the relationships between specific phenotypes and genomic changes is complicated by the fact that at the molecular level, most changes are associated not with selection but neutral processes. Distinguishing between the two is necessary for reliable interpretation of any potential signature of selection. In this respect, this thesis addresses three aspects of genome evolution: biasing factors for the interpretation of protein evolutionary rates, the adaptive significance of presence/absence variation (PAV) in the model plant A. thaliana and the evolution of alternative splicing alongside increasing organism complexity. A comparative genomics approach is used throughout, taking advantage of the increasing availability of high-throughput sequencing data. We show that (i) lineage-specific substitutions and the differential conservation of the edges of exons influence interpretations of protein evolutionary rate, (ii) that PAV in A. thaliana can be explained without invoking adaption, despite enrichments for PAV events in genes considered as positively selected, and (iii) that alternative splicing is amongst the strongest predictors of organism complexity, consistent with an adaptive role of transcript diversification in determining a genome’s functional information capacity. Taken together, we find that the signatures and targets of adaptation can either be masked by, or re-interpreted in the context of, non-adaptive processes and that with the increasing availability of high-throughput data, such considerations are of increasing relevance.