On the Robustness to Gene Tree Estimation Error (or lack thereof) of Coalescent-Based Species Tree Methods

Abstract The multispecies coalescent (MSC) model provides a natural framework for species tree estimation accounting for gene-tree conflicts. While a number of species tree methods under the MSC have been suggested and evaluated using simulation, their statistical properties remain poorly understood. Here we use mathematical analysis aided by computer simulation to examine the identifiability, consistency, and efficiency of different species tree methods in the case of three species and three sequences under the molecular clock. We consider four major species-tree methods including concatenation, two-step, independent-sites maximum likelihood (ISML) and maximum likelihood (ML). We develop approximations that predict that the probit transform of the species tree estimation error decreases linearly with the square root of the number of loci. Even in this simplest case major differences exist among the methods. Fulllikelihood methods are considerably more efficient than summary methods such as concatenation and two-step. They also provide estimates of important parameters such as species divergence times and ancestral population sizes while these parameters are not identifiable by summary methods. Our results highlight the need to improve the statistical efficiency of summary methods and the computational efficiency of full likelihood methods of species tree estimation.

Download Full-text

The Perfect Storm: Gene Tree Estimation Error, Incomplete Lineage Sorting, and Ancient Gene Flow Explain the Most Recalcitrant Ancient Angiosperm Clade, Malpighiales

Systematic Biology ◽

10.1093/sysbio/syaa083 ◽

2020 ◽

Author(s):

Liming Cai ◽

Zhenxiang Xi ◽

Emily Moriarty Lemmon ◽

Alan R Lemmon ◽

Austin Mast ◽

...

Keyword(s):

Gene Flow ◽

Incomplete Lineage Sorting ◽

Estimation Error ◽

Gene Tree ◽

Species Tree ◽

Flowering Plant ◽

Estimation Methods ◽

Lineage Sorting ◽

Tree Estimation ◽

Perfect Storm

Abstract The genomic revolution offers renewed hope of resolving rapid radiations in the Tree of Life. The development of the multispecies coalescent (MSC) model and improved gene tree estimation methods can better accommodate gene tree heterogeneity caused by incomplete lineage sorting (ILS) and gene tree estimation error stemming from the short internal branches. However, the relative influence of these factors in species tree inference is not well understood. Using anchored hybrid enrichment, we generated a data set including 423 single-copy loci from 64 taxa representing 39 families to infer the species tree of the flowering plant order Malpighiales. This order includes nine of the top ten most unstable nodes in angiosperms, which have been hypothesized to arise from the rapid radiation during the Cretaceous. Here, we show that coalescent-based methods do not resolve the backbone of Malpighiales and concatenation methods yield inconsistent estimations, providing evidence that gene tree heterogeneity is high in this clade. Despite high levels of ILS and gene tree estimation error, our simulations demonstrate that these two factors alone are insufficient to explain the lack of resolution in this order. To explore this further, we examined triplet frequencies among empirical gene trees and discovered some of them deviated significantly from those attributed to ILS and estimation error, suggesting gene flow as an additional and previously unappreciated phenomenon promoting gene tree variation in Malpighiales. Finally, we applied a novel method to quantify the relative contribution of these three primary sources of gene tree heterogeneity and demonstrated that ILS, gene tree estimation error, and gene flow contributed to 10.0%, 34.8%, and 21.4% of the variation, respectively. Together, our results suggest that a perfect storm of factors likely influence this lack of resolution, and further indicate that recalcitrant phylogenetic relationships like the backbone of Malpighiales may be better represented as phylogenetic networks. Thus, reducing such groups solely to existing models that adhere strictly to bifurcating trees greatly oversimplifies reality, and obscures our ability to more clearly discern the process of evolution.

Download Full-text

To include or not to include: The impact of gene filtering on species tree estimation methods

10.1101/149120 ◽

2017 ◽

Cited By ~ 1

Author(s):

Erin K. Molloy ◽

Tandy Warnow

Keyword(s):

Missing Data ◽

Incomplete Lineage Sorting ◽

Estimation Error ◽

Gene Tree ◽

Species Tree ◽

Species Trees ◽

Lineage Sorting ◽

Gene Filtering ◽

Tree Estimation ◽

The Impact

AbstractSpecies tree estimation from loci sampled from multiple genomes is now common, but is challenged by the heterogeneity across the genome due to multiple processes, such as gene duplication and loss, horizontal gene transfer, and incomplete lineage sorting. Although methods for estimating species trees have been developed that address gene tree heterogeneity due to incomplete lineage sorting, many of these methods operate by combining estimated gene trees and are hence vulnerable to gene tree quality. There is also the added concern that missing data, which is frequently encountered in genome-scale datasets, will impact species tree estimation.Our study addresses the impact of gene filtering on species trees inferred from multi-gene datasets. We address these questions using a large and heterogeneous collection of simulated datasets both with and without missing data. We compare several established coalescent-based methods (ASTRAL, ASTRID, MP-EST, and SVDquartets within PAUP*) as well as unpartitioned concatenation using maximum likelihood (RAxML).Our study shows that gene tree error and missing data impact all methods (and some methods degrade more than others), but the degree of incomplete lineage sorting and gene tree estimation error impacts the absolute and relative performance of methods as well as their response to gene filtering strategies. We find that filtering genes based on the degree of missing data is either neutral or else reduces the accuracy of all five methods examined, and so is not recommended. Filtering genes based on gene tree estimation error shows somewhat different trends. Under low levels of incomplete lineage sorting, removing genes with high gene tree estimation error can improve the accuracy of summary methods, but only if not too many genes are removed. Otherwise, filtering genes tends to increase error, especially under high levels of incomplete lineage sorting. Hence, while filtering genes based on missing data is not recommended, there are conditions under which removing high error gene trees can improve species tree estimation. This study provides insights into prior studies and suggests approaches for analyzing phylogenomic datasets.

Download Full-text

The Perfect Storm: Gene Tree Estimation Error, Incomplete Lineage Sorting, and Ancient Gene Flow Explain the Most Recalcitrant Ancient Angiosperm Clade, Malpighiales

10.1101/2020.05.26.112318 ◽

2020 ◽

Author(s):

Liming Cai ◽

Zhenxiang Xi ◽

Emily Moriarty Lemmon ◽

Alan R. Lemmon ◽

Austin Mast ◽

...

Keyword(s):

Gene Flow ◽

Incomplete Lineage Sorting ◽

Estimation Error ◽

Gene Tree ◽

Species Tree ◽

Flowering Plant ◽

Estimation Methods ◽

Lineage Sorting ◽

Tree Estimation ◽

Perfect Storm

ABSTRACTThe genomic revolution offers renewed hope of resolving rapid radiations in the Tree of Life. The development of the multispecies coalescent (MSC) model and improved gene tree estimation methods can better accommodate gene tree heterogeneity caused by incomplete lineage sorting (ILS) and gene tree estimation error stemming from the short internal branches. However, the relative influence of these factors in species tree inference is not well understood. Using anchored hybrid enrichment, we generated a data set including 423 single-copy loci from 64 taxa representing 39 families to infer the species tree of the flowering plant order Malpighiales. This order alone includes nine of the top ten most unstable nodes in angiosperms, and the recalcitrant relationships along the backbone of the order have been hypothesized to arise from the rapid radiation during the Cretaceous. Here, we show that coalescent-based methods do not resolve the backbone of Malpighiales and concatenation methods yield inconsistent estimations, providing evidence that gene tree heterogeneity is high in this clade. Despite high levels of ILS and gene tree estimation error, our simulations demonstrate that these two factors alone are insufficient to explain the lack of resolution in this order. To explore this further, we examined triplet frequencies among empirical gene trees and discovered some of them deviated significantly from those attributed to ILS and estimation error, suggesting gene flow as an additional and previously unappreciated phenomenon promoting gene tree variation in Malpighiales. Finally, we applied a novel method to quantify the relative contribution of these three primary sources of gene tree heterogeneity and demonstrated that ILS, gene tree estimation error, and gene flow contributed to 15%, 52%, and 32% of the variation, respectively. Together, our results suggest that a perfect storm of factors likely influence this lack of resolution, and further indicate that recalcitrant phylogenetic relationships like the backbone of Malpighiales may be better represented as phylogenetic networks. Thus, reducing such groups solely to existing models that adhere strictly to bifurcating trees greatly oversimplifies reality, and obscures our ability to more clearly discern the process of evolution.

Download Full-text

QT-GILD: Quartet Based Gene Tree Imputation Using Deep Learning Improves Phylogenomic Analyses Despite Missing Data

10.1101/2021.11.03.467204 ◽

2021 ◽

Author(s):

Sazan Mahbub ◽

Shashata Sawmya ◽

Arpita Saha ◽

Rezwana Reaz ◽

M. Sohel Rahman ◽

...

Keyword(s):

Deep Learning ◽

Missing Data ◽

Language Processing ◽

Estimation Error ◽

Gene Tree ◽

Experimental Studies ◽

Species Tree ◽

Dramatic Improvement ◽

Gene Trees ◽

Tree Estimation

Species tree estimation is frequently based on phylogenomic approaches that use multiple genes from throughout the genome. However, for a combination of reasons (ranging from sampling biases to more biological causes, as in gene birth and loss), gene trees are often incomplete, meaning that not all species of interest have a common set of genes. Incomplete gene trees can potentially impact the accuracy of phylogenomic inference. We, for the first time, introduce the problem of imputing the quartet distribution induced by a set of incomplete gene trees, which involves adding the missing quartets back to the quartet distribution. We present QT-GILD, an automated and specially tailored unsupervised deep learning technique, accompanied by cues from natural language processing (NLP), which learns the quartet distribution in a given set of incomplete gene trees and generates a complete set of quartets accordingly. QT-GILD is a general-purpose technique needing no explicit modeling of the subject system or reasons for missing data or gene tree heterogeneity. Experimental studies on a collection of simulated and empirical data sets suggest that QT-GILD can effectively impute the quartet distribution, which results in a dramatic improvement in the species tree accuracy. Remarkably, QT-GILD not only imputes the missing quartets but it can also account for gene tree estimation error. Therefore, QT-GILD advances the state-of-the-art in species tree estimation from gene trees in the face of missing data. QT-GILD is freely available in open source form at https://github.com/pythonLoader/QT-GILD .

Download Full-text

Detecting the anomaly zone in species trees and evidence for a misleading signal in higher-level skink phylogeny (Squamata: Scincidae).

10.1101/012096 ◽

2014 ◽

Cited By ~ 3

Author(s):

Charles W Linkem ◽

Vladimir N. Minin ◽

Adam D Leache

Keyword(s):

Gene Tree ◽

Species Tree ◽

Ancestral Population ◽

Real Problem ◽

Species Trees ◽

Phylogenetic Studies ◽

New Information ◽

Population Sizes ◽

Tree Methods ◽

Tree Estimation

The anomaly zone presents a major challenge to the accurate resolution of many parts of the Tree of Life. The anomaly zone is defined by the presence of a gene tree topology that is more probable than the true species tree. This discrepancy can result from consecutive rapid speciation events in the species tree. Similar to the problem of long-branch attraction, including more data (loci) will only reinforce the support for the incorrect species tree. Empirical phylogenetic studies often implement coalescent based species tree methods to avoid the anomaly zone, but to this point these studies have not had a method for providing any direct evidence that the species tree is actually in the anomaly zone. In this study, we use 16 species of lizards in the family Scincidae to investigate whether nodes that are difficult to resolve are located within the anomaly zone. We analyze new phylogenomic data (429 loci), using both concatenation and coalescent based species tree estimation, to locate conflicting topological signal. We then use the unifying principle of the anomaly zone, together with estimates of ancestral population sizes and species persistence times, to determine whether the observed phylogenetic conflict is a result of the anomaly zone. We identify at least three regions of the Scindidae phylogeny that provide demographic signatures consistent with the anomaly zone, and this new information helps reconcile the phylogenetic conflict in previously published studies on these lizards. The anomaly zone presents a real problem in phylogenetics, and our new framework for identifying anomalous relationships will help empiricists leverage their resources appropriately for overcoming this challenge.

Download Full-text

Phylogenetic conflicts, combinability, and deep phylogenomics in plants

10.1101/371930 ◽

2018 ◽

Cited By ~ 1

Author(s):

Stephen A. Smith ◽

Nathanael Walker-Hale ◽

Joseph F. Walker ◽

Joseph W. Brown

Keyword(s):

Phylogenetic Relationships ◽

Phylogenetic Signal ◽

Gene Tree ◽

Species Tree ◽

Gene Trees ◽

Data Filtering ◽

Tree Inference ◽

Tree Methods ◽

Inference Methods ◽

Species Tree Inference

AbstractStudies have demonstrated that pervasive gene tree conflict underlies several important phylogenetic relationships where different species tree methods produce conflicting results. Here, we present a means of dissecting the phylogenetic signal for alternative resolutions within a dataset in order to resolve recalcitrant relationships and, importantly, identify what the dataset is unable to resolve. These procedures extend upon methods for isolating conflict and concordance involving specific candidate relationships and can be used to identify systematic error and disambiguate sources of conflict among species tree inference methods. We demonstrate these on a large phylogenomic plant dataset. Our results support the placement of Amborella as sister to the remaining extant angiosperms, Gnetales as sister to pines, and the monophyly of extant gymnosperms. Several other contentious relationships, including the resolution of relationships within the bryophytes and the eudicots, remain uncertain given the low number of supporting gene trees. To address whether concatenation of filtered genes amplified phylogenetic signal for relationships, we implemented a combinatorial heuristic to test combinability of genes. We found that nested conflicts limited the ability of data filtering methods to fully ameliorate conflicting signal amongst gene trees. These analyses confirmed that the underlying conflicting signal does not support broad concatenation of genes. Our approach provides a means of dissecting a specific dataset to address deep phylogenetic relationships while also identifying the inferential boundaries of the dataset.

Download Full-text

Phylogenomic terraces: presence and implication in species tree estimation from gene trees

10.1101/2020.04.19.048843 ◽

2020 ◽

Author(s):

Ishrat Tanzila Farah ◽

Md Muktadirul Islam ◽

Kazi Tasnim Zinat ◽

Atif Hasan Rahman ◽

Md Shamsuzzoha Bayzid

Keyword(s):

Incomplete Lineage Sorting ◽

Gene Tree ◽

Species Tree ◽

Gene Trees ◽

Species Trees ◽

Lineage Sorting ◽

Deep Coalescence ◽

Tree Estimation ◽

Tree Space ◽

Multiple Species

AbstractSpecies tree estimation from multi-locus dataset is extremely challenging, especially in the presence of gene tree heterogeneity across the genome due to incomplete lineage sorting (ILS). Summary methods have been developed which estimate gene trees and then combine the gene trees to estimate a species tree by optimizing various optimization scores. In this study, we have formalized the concept of “phylogenomic terraces” in the species tree space, where multiple species trees with distinct topologies may have exactly the same optimization score (quartet score, extra lineage score, etc.) with respect to a collection of gene trees. We investigated the presence and implication of terraces in species tree estimation from multi-locus data by taking ILS into account. We analyzed two of the most popular ILS-aware optimization criteria: maximize quartet consistency (MQC) and minimize deep coalescence (MDC). Methods based on MQC are provably statistically consistent, whereas MDC is not a consistent criterion for species tree estimation. Our experiments, on a collection of dataset simulated under ILS, indicate that MDC-based methods may achieve competitive or identical quartet consistency score as MQC but could be significantly worse than MQC in terms of tree accuracy – demonstrating the presence and affect of phylogenomic terraces. This is the first known study that formalizes the concept of phylogenomic terraces in the context of species tree estimation from multi-locus data, and reports the presence and implications of terraces in species tree estimation under ILS.

Download Full-text

Whole-Genome Analyses Resolve the Phylogeny of Flightless Birds (Palaeognathae) in the Presence of an Empirical Anomaly Zone

Systematic Biology ◽

10.1093/sysbio/syz019 ◽

2019 ◽

Vol 68 (6) ◽

pp. 937-955 ◽

Cited By ~ 21

Author(s):

Alison Cloutier ◽

Timothy B Sackton ◽

Phil Grayson ◽

Michele Clamp ◽

Allan J Baker ◽

...

Keyword(s):

Sequence Data ◽

Incomplete Lineage Sorting ◽

Estimation Error ◽

Gene Tree ◽

Species Tree ◽

Tree Topology ◽

Whole Genome ◽

Gene Trees ◽

Lineage Sorting ◽

Genome Analyses

Abstract Palaeognathae represent one of the two basal lineages in modern birds, and comprise the volant (flighted) tinamous and the flightless ratites. Resolving palaeognath phylogenetic relationships has historically proved difficult, and short internal branches separating major palaeognath lineages in previous molecular phylogenies suggest that extensive incomplete lineage sorting (ILS) might have accompanied a rapid ancient divergence. Here, we investigate palaeognath relationships using genome-wide data sets of three types of noncoding nuclear markers, together totaling 20,850 loci and over 41 million base pairs of aligned sequence data. We recover a fully resolved topology placing rheas as the sister to kiwi and emu + cassowary that is congruent across marker types for two species tree methods (MP-EST and ASTRAL-II). This topology is corroborated by patterns of insertions for 4274 CR1 retroelements identified from multispecies whole-genome screening, and is robustly supported by phylogenomic subsampling analyses, with MP-EST demonstrating particularly consistent performance across subsampling replicates as compared to ASTRAL. In contrast, analyses of concatenated data supermatrices recover rheas as the sister to all other nonostrich palaeognaths, an alternative that lacks retroelement support and shows inconsistent behavior under subsampling approaches. While statistically supporting the species tree topology, conflicting patterns of retroelement insertions also occur and imply high amounts of ILS across short successive internal branches, consistent with observed patterns of gene tree heterogeneity. Coalescent simulations and topology tests indicate that the majority of observed topological incongruence among gene trees is consistent with coalescent variation rather than arising from gene tree estimation error alone, and estimated branch lengths for short successive internodes in the inferred species tree fall within the theoretical range encompassing the anomaly zone. Distributions of empirical gene trees confirm that the most common gene tree topology for each marker type differs from the species tree, signifying the existence of an empirical anomaly zone in palaeognaths.

Download Full-text

Optimizing Phylogenomics with Rapidly Evolving Long Exons: Comparison with Anchored Hybrid Enrichment and Ultraconserved Elements

Molecular Biology and Evolution ◽

10.1093/molbev/msz263 ◽

2019 ◽

Vol 37 (3) ◽

pp. 904-922 ◽

Cited By ~ 4

Author(s):

Benjamin R Karin ◽

Tony Gamble ◽

Todd R Jackman

Keyword(s):

Evolutionary Rate ◽

Estimation Error ◽

Model Misspecification ◽

Gene Tree ◽

Species Tree ◽

Gene Trees ◽

Data Set ◽

Ultraconserved Elements ◽

Trade Offs ◽

Anchored Hybrid Enrichment

Abstract Marker selection has emerged as an important component of phylogenomic study design due to rising concerns of the effects of gene tree estimation error, model misspecification, and data-type differences. Researchers must balance various trade-offs associated with locus length and evolutionary rate among other factors. The most commonly used reduced representation data sets for phylogenomics are ultraconserved elements (UCEs) and Anchored Hybrid Enrichment (AHE). Here, we introduce Rapidly Evolving Long Exon Capture (RELEC), a new set of loci that targets single exons that are both rapidly evolving (evolutionary rate faster than RAG1) and relatively long in length (>1,500 bp), while at the same time avoiding paralogy issues across amniotes. We compare the RELEC data set to UCEs and AHE in squamate reptiles by aligning and analyzing orthologous sequences from 17 squamate genomes, composed of 10 snakes and 7 lizards. The RELEC data set (179 loci) outperforms AHE and UCEs by maximizing per-locus genetic variation while maintaining presence and orthology across a range of evolutionary scales. RELEC markers show higher phylogenetic informativeness than UCE and AHE loci, and RELEC gene trees show greater similarity to the species tree than AHE or UCE gene trees. Furthermore, with fewer loci, RELEC remains computationally tractable for full Bayesian coalescent species tree analyses. We contrast RELEC to and discuss important aspects of comparable methods, and demonstrate how RELEC may be the most effective set of loci for resolving difficult nodes and rapid radiations. We provide several resources for capturing or extracting RELEC loci from other amniote groups.

Download Full-text