PoMoSelect: Disentangling Modes of Selection

Molecular phylogenetics has neglected polymorphisms within present and ancestral populations for a long time. Alternative models accounting for multi-individual data have nevertheless been proposed and are known as polymorphism-aware phylogenetic models (PoMo). PoMo adds a new layer of complexity to the standard nucleotide substitution models by accounting for the population-level (so far, genetic drift and mutations) processes to describe the evolutionary process. To do so, PoMo expands the standard substitution models to include polymorphic states. We have previously shown that PoMo accounts for incomplete lineage sorting (ILS), and improve the estimation of species tree inference.
For this project, we will develop an approach that accounts for balancing selection called PoMoSelect. Genetic drift removes polymorphism from populations over time, with the rate of polymorphism loss being accelerated when species experience strong reductions in population size. Adaptive forces that maintain genetic variation in populations, or balancing selection, might counteract this process. PoMo is a mutation-selection model which we will use disentangle balancing selection from directional selection as well as mutational effects, fixation biases, and demographic effects. Furthermore, PoMo combination of polymorphism with divergence data allows it to model short as well as long-term balancing selection. We will introduce a new mechanistic parameters to quantify the strength of balancing selection on a loci. For the inference of the new parameters we will develop a new Bayesian framework and software package PoMoSelect. Together with our collaborators, we will analyse DNA sequences of African hunter-gather and farmer populations as well as great ape species to understand the role of blood parasites on the alpha and beta globin cluster (thalasiamias). Finally, we will apply our new methodology genome-wide to study the extent and patterns of balancing selection.

Molecular phylogenetics has neglected polymorphisms within present and ancestral populations for a long time. Alternative models accounting for multi-individual data have nevertheless been proposed and are known as polymorphism-aware phylogenetic models (PoMo). PoMo adds a new layer of complexity to the standard nucleotide substitution models by accounting for the population-level (so far, genetic drift and mutations) processes to describe the evolutionary process. To do so, PoMo expands the standard substitution models to include polymorphic states. We have previously shown that PoMo accounts for incomplete lineage sorting (ILS), and improve the estimation of species tree inference.
For this project, we will develop an approach that accounts for balancing selection called PoMoSelect. Genetic drift removes polymorphism from populations over time, with the rate of polymorphism loss being accelerated when species experience strong reductions in population size. Adaptive forces that maintain genetic variation in populations, or balancing selection, might counteract this process. PoMo is a mutation-selection model which we will use disentangle balancing selection from directional selection as well as mutational effects, fixation biases, and demographic effects. Furthermore, PoMo combination of polymorphism with divergence data allows it to model short as well as long-term balancing selection. We will introduce a new mechanistic parameters to quantify the strength of balancing selection on a loci. For the inference of the new parameters we will develop a new Bayesian framework and software package PoMoSelect. Together with our collaborators, we will analyse DNA sequences of African hunter-gather and farmer populations as well as great ape species to understand the role of blood parasites on the alpha and beta globin cluster (thalasiamias). Finally, we will apply our new methodology genome-wide to study the extent and patterns of balancing selection.