Expanded metabolic networks combined with accelerated protein evolution has enabled new cellular traits within yeast subphylum

Novel traits arise from genome evolution, but it remains ambiguous which evolutionary events modulate trait diversity in yeast subphylum. Here, we performed a multi-dimensional evolution analysis to investigate the general mechanism shaping yeast phenome diversity. We reconstructed genome-scale metabolic models for 332 yeast species to characterize versatile metabolic functions. Subsequent model simulation and selection analysis indicate that most genes are negatively selected across species, while the evolution rate of gene and codon are pathway- and function-dependent. Maping expanded and horizontal transferred genes onto pathways verify that the extended metabolic network through gene duplication and enzyme promiscuity are prominent mechanism for innovations in substrates utilization. We further find that the positive selection on multiple genes, converging on EMP pathway, contribute to the formation of more complex traits, i.e., thermotolerance and Crabtree-effect. Our findings testify that for phenotypic diversification, multi-dimensional evolution from both metabolic network structure and individual protein elements are crucial.