PhyloAcc-GT: A Bayesian method for inferring patterns of substitution rate shifts and associations with binary traits under gene tree discordance

Abstract

An important goal of evolutionary genomics is to identify genomic regions whose substitution rates differ among lineages. For example, genomic regions experiencing accelerated molecular evolution in some lineages may provide insight into links between genotype to phenotype. Several comparative genomics methods have been developed to identify genomic accelerations between species, including a Bayesian method called PhyloAcc, which models shifts in substitution rate in multiple target lineages on a phylogeny. However, few methods consider the possibility of discordance between the trees of individual loci and the species tree due to incomplete lineage sorting, which might cause false positives. Here we present PhyloAcc-GT, which extends PhyloAcc by modeling gene tree heterogeneity to detect rate shifts across genomic regions. Given a species tree, we adopt the multispecies coalescent model as the prior distribution of gene trees, use Markov chain Monte Carlo (MCMC) for inference, and design novel MCMC moves to sample gene trees efficiently. Through extensive simulations, we show that PhyloAcc-GT outperforms PhyloAcc and other methods in identifying target-lineage-specific accelerations and detecting complex patterns of rate shifts, and is robust to specification of population size parameters. We apply PhyloAcc-GT to two examples of convergent evolution: flightlessness in ratites and marine mammal adaptations. PhyloAcc-GT is usually more conservative than PhyloAcc in calling convergent rate shifts because it identifies more accelerations on ancestral than on terminal branches. In summary, PhyloAcc-GT is a useful tool to identify shifts in substitution rate associated with specific target lineages while accounting for incomplete lineage sorting.

Unifying approaches from statistical genetics and phylogenetics for mapping phenotypes in structured populations

“…In a series of papers, Hahn and colleagues [60, 96, 97, 185, 186] have pointed out that conducting comparative tests using the species tree can be misleading: if the genes underlying a phenotype of interest have different topologies than that of a species tree—due to incomplete lineage sorting (ILS) and other processes [187]—researchers may mistake similarity due to common ancestry for convergence. This phenomenon, known as hemiplasy [188] may be relatively common [185, 186], motivating the development of ILS-aware phylogenetic regresssion [97] and PhyloG2P methods [189].…”

“…One could potentially use an analogous approach to model phenotypic evolution along gene trees within a species (to inform the construction of eGRM, for example). This would improve inferences from both tree-based GWAS (sensu [58, 59]) and from emerging phylogenetic comparative approaches that consider gene trees rather than just the species trees [101, 102, 186] (such approaches are important as only using a single species tree may lead one to mistake similarity due to common ancestry for convergence [63, 187, 188]).…”

Repeated evolution of similar phenotypes: Integrating comparative methods with developmental pathways

“…, 2020 RERconverge Kowalczyk et al. , 2019 Coevol Lartillot & Poujol, 2011 PhyloAcc Hu et al , 2019 PhyloAcc-GT Yan et al. , 2022 Ancestral state reconstruction make.simmap/phytools R Revell, 2012 baseML or codeML PAML Yang, 2007 Selection tests (Branch/Clade Model) codeML PAML Yang, 2007 aBSREL HyPhy Smith et al.…”

“…Ignoring the mismatch between gene and species trees may result in incorrect estimates of substitution rates when mapping sequences from conflicting loci in the species tree ( Mendes et al , 2016 ). To overcome such a challenge, some programs consider gene tree heterogeneity in their approach ( Guerrero and Hahn, 2018 ; Yan et al , 2022 ). Despite the vast majority of models treating phenotypes as binary, there are some models that consider associations between genomic substitution rates and continuous phenotypes in the analyses implemented (see Kowalczyk et al , 2019 ).…”