Gene function predicts divergence across molecular levels

Genes can evolve through changes at their DNA, RNA, or protein levels. However, because these changes are measured using distinct and often incomparable metrics, their relative contributions to genic evolution remain unclear. Here, we address this challenge by developing a standardized framework for comparing evolutionary divergence in protein-coding sequences, multi-tissue expression profiles, and protein structures. Application of this approach to data from Drosophila, rodents, primates, and Arabidopsis reveals substantial variation in divergence patterns across taxa, with only sequence divergence matching expectations based on evolutionary theory. Moreover, sequences evolve slowest and protein structures fastest in all taxa, consistent with the idea that higher levels of biological organization are closer proxies for the functions on which selection acts. Yet, weak correlations among divergence measures suggest that targets of selection vary across genes, perhaps depending on their functions. Indeed, few genes exhibit similar evolutionary patterns across all three levels, and different functions are enriched in genes with low or high divergence in protein-coding sequences, gene expression profiles, and protein structures. Together, these findings support the hypothesis that evolutionary targets of genes depend on their functions, shedding light on how selection shapes different levels of biological organization across taxa.