PoMo-cod: a polymorphism-aware phylogenetic codon model

A major goal in evolutionary biology is to understand the forces that operate in the genomic sequences and are responsible for the adaptation of species to different environments. Codon models are one of the main tools used to infer selection on protein-coding genes. These have been popularized in comparative genomic studies by their extensive use in genome-wide scans of diversifying selection. However, models of codon evolution have significant limitations that are increasingly being recognized. The main one being that current codon models make simplistic assumptions, such as ignoring species demography and nucleotide usage bias. This project offers a new polymorphism-aware model, PoMo-cod, to detect signatures of natural selection acting on protein-coding sequences. PoMo-cod will address codon evolution in a unique way by properly reconciling the neutral and adaptive population-level processes by which coding sequences evolve. More importantly, PoMo-cod will allow us to tell apart the sole action of natural selection from known confounding forces (e.g., fluctuating demography and GC-biased gene conversion), ultimately producing more accurate genome-wide maps of diversifying evolving genes. Selection maps have impactful applications in other fields. They permit developing species-specific conservation strategies to mitigate the anthropogenic action on biodiversity or characterize the genome functionally, which has direct implications for medical research.

How living organisms adapt to their environment is one of the big questions in biology. This project aimed to improve how scientists study this process by developing new tools to understand how natural selection shapes the genome over time. By the end of the project, we delivered a set of phylogenetic models and methods that make it possible to study evolution across both short and long time scales in a more unified and realistic way. One of the main outcomes of the project is a new modelling framework describing how genetic differences can be maintained, rapidly eliminated, or promoted within and between species. This type of evolution, known as diversifying selection, plays a key role in adaptation. By connecting evolutionary processes acting at different time scales, this framework represents an important step forward compared to previous approaches. We expect it to have important implications for how scientists measure the effects of new mutations on the fitness of organisms. Another major achievement is the integration of changing population sizes into phylogenetic analyses. In nature, populations rarely remain stable: they grow, shrink, and sometimes disappear altogether. Many existing models ignore these fluctuations, which can distort the interpretation of genetic data. The methods developed in this project explicitly account for such changes using the concept of an eigenvalue population size, leading to more accurate reconstructions of evolutionary history. We also extended existing evolutionary models to better reflect species with very high genetic diversity, where the same mutations can occur repeatedly over time. This situation is common in organisms such as viruses and bacteria. By adapting earlier models, we made them suitable for studying adaptation in a much broader range of species than was previously possible. A further outcome of the project is a comprehensive set of Bayesian phylogenetic methods that has been made available to the scientific community. We also applied the new models and methods to real biological data, ranging from fast-evolving viruses to long-lived species such as great apes. These applications demonstrate the flexibility of the approach and show that the same mathematical framework can be used across very different forms of life. Overall, the project provides new ways to better understand the rules that govern life. We anticipate that the knowledge produced will support biodiversity conservation by identifying genomic regions with adaptive potential in changing environments, and will also contribute to medical research by improving our understanding of how pathogens evolve.