Error rates in Q_ST--F_ST comparisons depend on genetic architecture and estimation procedures DOI Creative Commons

J. Liu,

Michael D. Edge

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Ноя. 2, 2024

Abstract Genetic and phenotypic variation among populations is one of the fundamental subjects evolutionary genetics. One question that arises often in data on natural whether differentiation a particular trait might be caused part by selection. For past several decades, researchers have used Q ST – F approaches to compare amount or more traits (measured statistic ) with genome-wide genetic variants ). Theory says under neutrality, should approximately equal expectation, so values much larger than are consistent local adaptation driving subpopulations’ apart, smaller stabilizing selection similar optima. At same time, investigators differed their definitions (such as “ratio averages” vs. “average ratios” versions variance components . Here, we show these details matter. Different different interpretations terms coalescence comparing incompatible statistics can lead elevated type I error rates, some choices leading rates near when nominal rate 5%. We conduct simulations varying architectures forms population structure how they affect distribution When many loci influence trait, our support procedures grounded coalescent-based framework for neutral phenotytpic differentiation.

Язык: Английский

The meaning and measure of concordance factors in phylogenomics DOI Creative Commons
Robert Lanfear, Matthew W. Hahn

Molecular Biology and Evolution, Год журнала: 2024, Номер 41(11)

Опубликована: Окт. 17, 2024

Abstract As phylogenomic datasets have grown in size, researchers developed new ways to measure biological variation and assess statistical support for specific branches. Larger more sites loci therefore less sampling variance. While we can accurately the mean signal these datasets, lower variance is often reflected uniformly high measures of branch support—such as bootstrap posterior probability—limiting their utility. also revealed substantial topologies found across individual loci, such that single species tree inferred by most phylogenetic methods represents a limited summary data many purposes. In contrast support, degree underlying topological among should be approximately constant regardless size dataset. “Concordance factors” (CFs) similar statistics become increasingly important tools phylogenetics. this review, explain why CFs thought descriptors rather than argue they provide information about predictive power not contained support. We review growing suite measuring concordance, compare them common framework reveals interrelationships, demonstrate how calculate using an example from birds. discuss might change future move beyond estimating “tree life” toward myriad evolutionary histories genomic variation.

Язык: Английский

Процитировано

4

Multivariate Trait Evolution: Models for the Evolution of the Quantitative Genetic G-Matrix on Phylogenies. DOI Creative Commons
Simon P. Blomberg, Michelle Muniz,

Mai Bui

и другие.

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Окт. 29, 2024

Abstract Genetic covariance matrices (G-matrices) are a key focus for research and predictions from quantitative genetic evolutionary models of multiple traits. There is consensus among geneticists that the G-matrix can evolve through “deep” time. Yet, evolution conspicuously lacking. In contrast, field macroevolution has several stochastic univariate traits evolving on phylogenies. However, despite much into multivariate phylogenetic comparative methods, analytical how trait might phylogenies have not been considered. Here we show three phylogenies, based Lie group theory, Riemannian geometry differential (diffusion) equations, be combined to unify genetics macroevolutionary theory in coherent mathematical framework. The provide basis understanding G-matrices fit data via simulation using Approximate Bayesian Computation. Such used generate test hypotheses about variances covariances, together with themselves, these vary across phylogeny. This unification an important advance study phenotypes, allowing construction synthetic species over Lay Summary We unite Quantitative Genetics, major microevolution, macroevolution. To do this, allow component matrix additive covariances (the G-matrix) along trees. because assumed constant genetics, but it recognised evolves timescales (in “deep time”). Uniting Genetics allows more complete description Darwin’s evolution, further testing hypotheses.

Язык: Английский

Процитировано

2

Convergent expansions of keystone gene families drive metabolic innovation in a major eukaryotic clade DOI Creative Commons
Kyle David, Joshua G. Schraiber, Johnathan G. Crandall

и другие.

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Июль 23, 2024

Many remarkable innovations have repeatedly occurred across vast evolutionary distances. When convergent traits emerge on the tree of life, they are sometimes driven by same underlying gene families, while other times many different families involved. Conversely, a family may be recruited for single trait or traits. To understand general rules governing convergence at both genomic and phenotypic levels, we systematically tested associations between 56 binary metabolic count in 14,710 from 993 species Saccharomycotina yeasts. Using recently developed phylogenetic approach that reduces spurious correlations, discovered expansion contraction was significantly linked to gain loss 45/56 (80%) While 601/746 (81%) significant were associated with only one trait, also identified several 'keystone' up 13/56 (23%) all These results indicate yeasts governed narrow set major genetic elements mechanisms.

Язык: Английский

Процитировано

1

Error rates in Q_ST--F_ST comparisons depend on genetic architecture and estimation procedures DOI Creative Commons

J. Liu,

Michael D. Edge

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Ноя. 2, 2024

Abstract Genetic and phenotypic variation among populations is one of the fundamental subjects evolutionary genetics. One question that arises often in data on natural whether differentiation a particular trait might be caused part by selection. For past several decades, researchers have used Q ST – F approaches to compare amount or more traits (measured statistic ) with genome-wide genetic variants ). Theory says under neutrality, should approximately equal expectation, so values much larger than are consistent local adaptation driving subpopulations’ apart, smaller stabilizing selection similar optima. At same time, investigators differed their definitions (such as “ratio averages” vs. “average ratios” versions variance components . Here, we show these details matter. Different different interpretations terms coalescence comparing incompatible statistics can lead elevated type I error rates, some choices leading rates near when nominal rate 5%. We conduct simulations varying architectures forms population structure how they affect distribution When many loci influence trait, our support procedures grounded coalescent-based framework for neutral phenotytpic differentiation.

Язык: Английский

Процитировано

0