MIPhy: identify and quantify rapidly evolving members of large gene families

Abstract

After transitioning to a new environment, species often exhibit rapid phenotypic innovation. One of the fastest mechanisms for this is duplication followed by specialization of existing genes. When this happens to a member of a gene family, it tends to leave a detectable phylogenetic signature of lineage-specific expansions and contractions. These can be identified by analyzing the gene family across several species and identifying patterns of gene duplication and loss that do not correlate with the known relationships between those species. This signature, termed phylogenetic instability, has been previously linked to adaptations that change the way an organism samples and responds to its environment; conversely, low phylogenetic instability has been previously linked to proteins with endogenous functions. With the increase in genome-level data, there is a need to identify and quantify phylogenetic instability. Here, we present Minimizing Instability in Phylogenetics (MIPhy), a tool that solves this problem by quantifying the incongruence of a gene’s evolutionary history. The motivation behind MIPhy was to produce a tool to aid in interpreting phylogenetic trees. It can predict which members of a gene family are under adaptive evolution, working only from a gene tree and the relationship between the species under consideration. While it does not conduct any estimation of positive selection—which is the typical indication of adaptive evolution—the results tend to agree. We demonstrate the usefulness of MIPhy by accurately predicting which members of the mammalian cytochrome P450 gene superfamily metabolize xenobiotics and which metabolize endogenous compounds. Our predictions correlate very well with known substrate specificities of the human enzymes. We also analyze the Caenorhabditis collagen gene family and use MIPhy to predict genes that produce an observable phenotype when knocked down in C. elegans, and show that our predictions correlate well with existing knowledge. The software can be downloaded and installed from and is also available as an online web tool at .

Two novel loci underlie natural differences in Caenorhabditis elegans abamectin responses

“…The VL and VC QTL contain several gene families, including ugt and cyp genes, which have been implicated in resistance previously [ 58 , 59 ]. Gene families often evolve rapidly in response to the environment and are good candidates to confer resistance [ 60 ]. Because comparisons of all gene families or ortholog groups in the VL and VC QTL is difficult, we focused on ortholog groups for a subset of previously described candidate gene families ( ugt and cyp ).…”

Species tree disequilibrium positively misleads models of gene family evolution

“…We determined phylogenetic instability for all nine OR gene subfamilies in Phyllostomids (figure 2) using MIPhy (Minimizing Instability in Phylogenetics) (Curran et al 2018) and gene and species trees from Hayden et al (2014). MIPhy uses patterns of duplications and losses in gene trees to identify groups of genes that are under selection based on deviations from expected baseline patterns.…”

“…To quantify the phylogenetic instability for each of the nine OR gene families, we used the MIPhy online web tool with default weights (Curran et al 2018). The species tree and the list of OR genes corresponding to each species were compiled into an information file compatible with MIPhy.…”

“…Given the potential confounding effects of tree heterogeneity on birth-death equilibrium, we used simulations to determine if such an effect was present in models integrating the birth-death process with the phylogeny. We also used two approaches, both based on a combination of gene and species trees, to infer adaptation in olfactory receptors to the new frugivorous diet: Poisson mixed models that estimate rates of birth and death (Sackton et al 2017), and phylogenetic instability analyses that uses gene-tree/speciestree reconciliation (Curran et al 2018). We hypothesize that one or more OR gene families involved in shifts to frugivory have expanded in number relative to other OR gene families across bats.…”

Two novel loci underlie natural differences inCaenorhabditis elegansmacrocyclic lactone responses

“…The VL and VC QTL contain several gene families, including ugt and cyp genes, which have been implicated in resistance previously[53,54]. Gene families often evolve rapidly in response to the environment and are good candidates to confer resistance[55]. Because comparisons of all gene families or ortholog groups in the VL and VC QTL is difficult, we focused on ortholog groups for a subset of previously described candidate gene families (ugt and cyp).…”