Terraces in phylogenetic tree space

Research output: Contribution to journalArticle

60 Citations (Scopus)

Abstract

A key step in assembling the tree of life is the construction of species-rich phylogenies from multilocus - but often incomplete - sequence data sets. We describe previously unknown structure in the landscape of solutions to the tree reconstruction problem, comprising sometimes vast "terraces" of trees with identical quality, arranged on islands of phylogenetically similar trees. Phylogenetic ambiguity within a terrace can be characterized efficiently and then ameliorated by new algorithms for obtaining a terrace's maximum-agreement subtree or by identifying the smallest set of new targets for additional sequencing. Algorithms to find optimal trees or estimate Bayesian posterior tree distributions may need to navigate strategically in the neighborhood of large terraces in tree space.

Original languageEnglish (US)
Pages (from-to)448-450
Number of pages3
JournalScience
Volume333
Issue number6041
DOIs
StatePublished - Jul 22 2011

Fingerprint

Phylogeny
Islands
Datasets

ASJC Scopus subject areas

  • General

Cite this

Terraces in phylogenetic tree space. / Sanderson, Michael; Mcmahon, Michelle M; Steel, Mike.

In: Science, Vol. 333, No. 6041, 22.07.2011, p. 448-450.

Research output: Contribution to journalArticle

Sanderson, Michael ; Mcmahon, Michelle M ; Steel, Mike. / Terraces in phylogenetic tree space. In: Science. 2011 ; Vol. 333, No. 6041. pp. 448-450.
@article{b2a088fe00e34834b4aea0a180b0b54b,
title = "Terraces in phylogenetic tree space",
abstract = "A key step in assembling the tree of life is the construction of species-rich phylogenies from multilocus - but often incomplete - sequence data sets. We describe previously unknown structure in the landscape of solutions to the tree reconstruction problem, comprising sometimes vast {"}terraces{"} of trees with identical quality, arranged on islands of phylogenetically similar trees. Phylogenetic ambiguity within a terrace can be characterized efficiently and then ameliorated by new algorithms for obtaining a terrace's maximum-agreement subtree or by identifying the smallest set of new targets for additional sequencing. Algorithms to find optimal trees or estimate Bayesian posterior tree distributions may need to navigate strategically in the neighborhood of large terraces in tree space.",
author = "Michael Sanderson and Mcmahon, {Michelle M} and Mike Steel",
year = "2011",
month = "7",
day = "22",
doi = "10.1126/science.1206357",
language = "English (US)",
volume = "333",
pages = "448--450",
journal = "Science",
issn = "0036-8075",
publisher = "American Association for the Advancement of Science",
number = "6041",

}

TY - JOUR

T1 - Terraces in phylogenetic tree space

AU - Sanderson, Michael

AU - Mcmahon, Michelle M

AU - Steel, Mike

PY - 2011/7/22

Y1 - 2011/7/22

N2 - A key step in assembling the tree of life is the construction of species-rich phylogenies from multilocus - but often incomplete - sequence data sets. We describe previously unknown structure in the landscape of solutions to the tree reconstruction problem, comprising sometimes vast "terraces" of trees with identical quality, arranged on islands of phylogenetically similar trees. Phylogenetic ambiguity within a terrace can be characterized efficiently and then ameliorated by new algorithms for obtaining a terrace's maximum-agreement subtree or by identifying the smallest set of new targets for additional sequencing. Algorithms to find optimal trees or estimate Bayesian posterior tree distributions may need to navigate strategically in the neighborhood of large terraces in tree space.

AB - A key step in assembling the tree of life is the construction of species-rich phylogenies from multilocus - but often incomplete - sequence data sets. We describe previously unknown structure in the landscape of solutions to the tree reconstruction problem, comprising sometimes vast "terraces" of trees with identical quality, arranged on islands of phylogenetically similar trees. Phylogenetic ambiguity within a terrace can be characterized efficiently and then ameliorated by new algorithms for obtaining a terrace's maximum-agreement subtree or by identifying the smallest set of new targets for additional sequencing. Algorithms to find optimal trees or estimate Bayesian posterior tree distributions may need to navigate strategically in the neighborhood of large terraces in tree space.

UR - http://www.scopus.com/inward/record.url?scp=79960670940&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79960670940&partnerID=8YFLogxK

U2 - 10.1126/science.1206357

DO - 10.1126/science.1206357

M3 - Article

VL - 333

SP - 448

EP - 450

JO - Science

JF - Science

SN - 0036-8075

IS - 6041

ER -