Edge Rewrite
everything Source code View original
// HTMLRewriter · presentation

This page was redesigned at the edge.

Cloudflare fetched the original article and streamed it through HTMLRewriter to apply an entirely new visual system without rebuilding the source page.

// request.cf · coarse context

A page that knows where it met you.

Only coarse request metadata is shown. This demo does not display or persist visitor IP addresses.

Country
US
Cloudflare location
CMH
Connection
HTTP/2
Language
Not provided

Ray ID: a22541f3ebbc910d

Jump to content

Draft:StarBEAST

From Wikipedia, the free encyclopedia
StarBEAST
DevelopersJoseph Heled, Alexei Drummond, Huw Ogilvie, Remco Bouckaert, Jordan Douglas and others
Release2010; 16 years ago (2010)
Stable release
StarBeast3
Written inJava
Operating systemCross-platform
TypeBioinformatics, Phylogenetics
LicenseLGPL
Websitewww.beast2.org

StarBEAST (originally styled *BEAST and pronounced "star beast") is a Bayesian Markov chain Monte Carlo (MCMC) method and software package for the joint estimation of species trees, gene trees, divergence times and ancestral effective population sizes from multilocus molecular sequence data, under the multispecies coalescent (MSC) model. It is implemented as a component of the BEAST platform for Bayesian evolutionary analysis. By modelling the coalescent process within a species tree, StarBEAST accounts for incomplete lineage sorting and the resulting discordance between individual gene trees, a source of error not addressed by the simple concatenation of loci.[1] It is one of the standard full-likelihood implementations of the multispecies coalescent and is described as such in independent reviews of species-tree inference.[2][3]

Since its introduction, the method has been reimplemented and extended through three major versions: the original *BEAST (2010), StarBEAST2 (2017) and StarBeast3 (2022), each substantially improving computational performance and adding modelling features.

Background

[edit]

Species that have diverged recently often share ancestral genetic polymorphisms that sort independently in different parts of the genome, so that the evolutionary tree estimated from any single gene may differ from the true species tree. This phenomenon, known as incomplete lineage sorting, means that a species tree cannot in general be recovered reliably by concatenating multiple loci and analysing them as a single alignment; concatenation can be statistically inconsistent under high levels of incomplete lineage sorting, converging on an incorrect tree with strong support.[4] The multispecies coalescent model addresses this by treating each gene tree as an outcome of the coalescent process running backwards in time within the branches of a shared species tree, with the rate of coalescence governed by the effective population size on each branch.[1] Because they remain consistent in this regime, full-likelihood multispecies-coalescent methods such as StarBEAST are used despite being computationally expensive relative to concatenation and summary methods.[2][4]

StarBEAST performs Bayesian inference of the species tree and its parameters by co-estimating the embedded gene trees, integrating over the uncertainty in each, rather than treating pre-estimated gene trees as fixed data.[1]

Versions

[edit]

*BEAST (2010)

[edit]

The original method, *BEAST, was described by Joseph Heled and Alexei Drummond in 2010 and implemented within BEAST 1. It provided a full Bayesian framework that jointly infers the species tree topology, divergence times, population sizes and gene trees from multiple genes sampled from multiple individuals across a set of closely related species. A birth–death or Yule prior is placed on the species tree, and a multispecies coalescent prior links the gene trees to it.[1]

StarBEAST2 (2017)

[edit]

StarBEAST2, published by Huw A. Ogilvie, Remco R. Bouckaert and Alexei J. Drummond in 2017, is a package for BEAST 2. It re-engineered the model and MCMC operators for greater efficiency, reporting speed-ups of roughly 13.5× and 13.8× on two empirical data sets and an average of about 33.1× across 30 simulated data sets relative to *BEAST. A key methodological addition was a set of MSC-aware relaxed molecular clock models, allowing accurate estimation of per-species substitution rates alongside the species tree. StarBEAST2 also analytically integrates over the ancestral effective population sizes by default, marginalizing these parameters rather than sampling them explicitly, which improves per-step performance.[5]

StarBeast3 (2022)

[edit]

StarBeast3, published by Jordan Douglas, Cinthy L. Jiménez-Silva and Remco Bouckaert in 2022, is a further BEAST 2 package designed for large data sets. It introduced new MCMC proposals and an adaptive, parallelised inference scheme in which conditionally independent gene trees and their site models can be updated across multiple threads. The authors reported that StarBeast3 is up to about one and a half orders of magnitude faster than StarBEAST2 and more than two orders of magnitude faster than *BEAST, depending on the data set and parameter, enabling analyses that combine hundreds of loci with relaxed-clock dating.[6]

Features

[edit]

Across its versions, StarBEAST provides:

  • Joint Bayesian estimation of the species tree, gene trees and divergence times under the multispecies coalescent.[1]
  • Estimation of ancestral effective population sizes on species-tree branches.[1]
  • Multispecies-coalescent relaxed-clock models for per-lineage substitution-rate estimation (from StarBEAST2 onward).[5]
  • Adaptive, parallelised MCMC for scaling to many loci (StarBeast3).[6]
  • Integration with the wider BEAST 2 ecosystem, including the BEAUti interface for building analyses and Tracer and DensiTree for inspecting results.[7]

Reception and use

[edit]

StarBEAST is one of the most widely used implementations of the multispecies coalescent for species-tree inference. In a review of Bayesian multispecies-coalescent methods, Jiao, Flouri and Yang describe *BEAST and BPP as "the two Bayesian programs implementing the model in common use".[3] Rannala, Edwards, Leaché and Yang likewise identify *BEAST as one of the standard full-likelihood programs for species-tree inference under the model in their review chapter in Phylogenetics in the Genomic Era.[2] Reviews of coalescent-based species-tree estimation discuss *BEAST as a fully Bayesian method that co-estimates the gene trees and the species tree, while noting that its computational cost limits the number of loci that can be analysed;[4] method-comparison studies use it as a reference point when evaluating faster summary and site-based approaches.[8]

The method is applied across evolutionary biology, including to genomic-scale target-capture datasets, for example in phylogenomic studies of Old World treefrogs and of Central African rain-forest plants.[9][10]

See also

[edit]

References

[edit]
  1. ^ a b c d e f Heled, Joseph; Drummond, Alexei J. (2010). "Bayesian Inference of Species Trees from Multilocus Data". Molecular Biology and Evolution. 27 (3): 570–580. doi:10.1093/molbev/msp274. PMID 19906723.
  2. ^ a b c Rannala, Bruce; Edwards, Scott V.; Leaché, Adam D.; Yang, Ziheng (2020). "The Multispecies Coalescent Model and Species Tree Inference". In Scornavacca, Céline; Delsuc, Frédéric; Galtier, Nicolas (eds.). Phylogenetics in the Genomic Era. pp. 3.3:1–3.3:21.
  3. ^ a b Jiao, Xiyun; Flouri, Tomáš; Yang, Ziheng (2021). "Multispecies coalescent and its applications to infer species phylogenies and cross-species gene flow". National Science Review. 8 (12) nwab127. doi:10.1093/nsr/nwab127.
  4. ^ a b c Liu, Liang; Wu, Shaoyuan; Yu, Lili (2015). "Coalescent methods for estimating species trees from phylogenomic data". Journal of Systematics and Evolution. 53 (5): 380–390. doi:10.1111/jse.12160.
  5. ^ a b Ogilvie, Huw A.; Bouckaert, Remco R.; Drummond, Alexei J. (2017). "StarBEAST2 Brings Faster Species Tree Inference and Accurate Estimates of Substitution Rates". Molecular Biology and Evolution. 34 (8): 2101–2114. doi:10.1093/molbev/msx126. PMC 5850801. PMID 28431121.
  6. ^ a b Douglas, Jordan; Jiménez-Silva, Cinthy L.; Bouckaert, Remco (2022). "StarBeast3: Adaptive Parallelized Bayesian Inference under the Multispecies Coalescent". Systematic Biology. 71 (4): 901–916. doi:10.1093/sysbio/syac010.
  7. ^ Bouckaert, Remco; Vaughan, Timothy G.; Barido-Sottani, Joëlle; Duchêne, Sebastián; Fourment, Mathieu; Gavryushkina, Alexandra; et al. (2019-04-08). "BEAST 2.5: An advanced software platform for Bayesian evolutionary analysis". PLOS Computational Biology. 15 (4) e1006650. Bibcode:2019PLSCB..15E6650B. doi:10.1371/journal.pcbi.1006650. PMC 6472827. PMID 30958812. S2CID 104294209.
  8. ^ Chou, Jed; Gupta, Ashu; Yaduvanshi, Shashank; Davidson, Ruth; Nute, Michael; Mirarab, Siavash; Warnow, Tandy (2015). "A comparative study of SVDquartets and other coalescent-based species tree estimation methods". BMC Genomics. 16 (Suppl 10): S2. doi:10.1186/1471-2164-16-S10-S2. PMC 4602346.
  9. ^ Chan, Kin Onn; Hutter, Carl R.; Wood, Perry L.; Grismer, L. Lee; Brown, Rafe M. (2020). "Target-capture phylogenomics provide insights on gene and species tree discordances in Old World treefrogs (Anura: Rhacophoridae)". Proceedings of the Royal Society B. 287 (1940) 20202102. doi:10.1098/rspb.2020.2102. PMID 33290680.
  10. ^ Helmstetter, Andrew J.; Béthune, Kevin; Kamdem, Narcisse G.; Sonké, Bonaventure; Couvreur, Thomas L. P. (2020). "Individualistic evolutionary responses of Central African rain forest plants to Pleistocene climatic fluctuations". Proceedings of the National Academy of Sciences. 117 (51): 32509–32518. doi:10.1073/pnas.2001018117. PMC 7768702.
[edit]

Category:Bioinformatics software Category:Phylogenetics software Category:Computational phylogenetics Category:Free science software

Retrieved from "https://en.wikipedia.org/w/index.php?title=Draft:StarBEAST&oldid=1366513896"