PoMo-cod: Integration von Demografie und Selektion in der Kodon-Evolution

PoMo-cod: Integrating demography and selection in codon evolutionA Comprehensive framework for evolutionary genomicsModelling sequence evolution continues to be an essential tool for understanding why, when, and how the evolutionary process occurs. Modelling the evolution of DNA sequences aids our understanding on how mutations are introduced and evolve along the genome. On the other hand, studying amino acid sequences can tell us how natural selection has shaped the history of a lineage, as changes in the protein affect the fitness of organisms. The evolution of coding sequences, in particular, offers a unique window into this intricate process as it sheds light on both DNA-level mutations and natural selection’s impact on proteins.Existing models of codon evolution (e.g., Goldman and Yang (1994); Muse and Gaut (1994)) provide us with a basic framework for testing the adaptive factors acting on the protein. They indicate where and when changes in selection pressures have occurred. Despite these contributions, these models only consider differences between species based on a single representative individual, ignoring variation observed within species/populations (polymorphisms). With recent advances in sequencing technology, together with the consistent reduction of associated costs, there is a need for tools that can describe adaptive forces within species as well as between them, incorporating the diversity found within a population, which would otherwise be neglected by the use of single-individual data.Furthermore, the current models could benefit from more realistic aspects, such as the ability to describe sophisticated mutation models, variable selection pressures, nucleotide usage biases and changing demography (changes in population size) – a key process determining the intensity of selection regimes. In response to these challenges, we present a new and more complex model, PoMo-cod (POlymorphism-aware phylogenetic MOdel for CODon evolution), adapted from polymorphism models of DNA evolution that were developed in our lab (De Maio et al., 2013; Borges et al., 2019). Our research aims to:• explicitly model positive diversifying selection, the type of selection that favours recurrent changes in the amino acid sequence. Positive diversifying selection is the main force driving adaptation. • incorporate varying demography, GC-biased gene conversion, and selection on synonymous sites in our model together with mutation and genetic drift. • apply our model to detect patterns of diversifying selection within the fruit fly and great ape datasets, which are characterised by distinct demographic and genomic attributes.By accounting for these confounded forces of demography, selection on synonymous sites, and GC-biased gene conversion, together with the inclusion of segregating sites (i.e. polymorphisms), we can create more accurate genome-wide maps of diversifying selection to understand the intricate process of species adaptation. Our software will harness novel computational methods and inferential strategies, and with additional time and resources, we will:• implement our software in both Bayesian and Maximum likelihood frameworks,• explore the application of cutting-edge Artificial Intelligence and Machine Learning to alleviate some of the computational burden associated with parameter-rich phylogenetic models.

ÖAW DOC

Zentrum für Biologische Wissenschaften