Re: [R-sig-phylo] Why no branch lengths on consensus trees?
Hi Scott, The reason for implementing only the consensus on the topology was after reading the appropriate chapter in Inferring Phylogenies. So I'm glad that Joe himself stepped in. I have added this reference in the help page (it was already cited in my book with respect to this issue). I think that the usage of branch lengths on (or from) a consensus tree depends on your perspective on phylogenetic estimation. If you are a frequentist, the goal of resampling the data is to give an idea of the accuracy of the estimates. So you might not use the bootstrap trees to estimate branch lengths; you already did that by ML (or maybe least squares) with the full data. If you are Bayesian, the trees sampled from an MCMC are here for estimation including of the branch lengths, so you use them to compute some sort of consensus topology as well as its branch lengths. So it makes sense that MrBayes can do a consensus tree with branch lengths. Cheers, Emmanuel Scott Chamberlain wrote on 22/11/2012 09:41: Dear Joe, Thanks for your feedback on this question. I will go read those pages you mentioned. Scott On Wed, Nov 21, 2012 at 5:19 PM, Joe Felsensteinj...@gs.washington.eduwrote: Daniel Barker wrote: What should branch lengths on a consensus tree represent? Scott Chamberlain had written: When making a consensus tree using ape::consensus the branch lengths are lost. Is there a way to not lose the branch lengths? Or to add them somehow to the consensus tree after making it. The issue of what branch lengths ought to be on a consensus tree is not simple. If we have three rooted trees: ((A:1,B:1):1,C:2); ((A:1,B:1):1,C:2); (A:2,(B:1,C:1):1); the consensus tree should be the first tree, but what branch length should be used for (say) the branch ancestral to the AB clade? 1? 0.667? The minute you open this can of worms it becomes clear that the answer depends on what you want that number to convey and what interpretations your audience will tend to draw from the number. There is no obvious answer. So this is not a mere technical computing question. By the way, in my 2004 book, you will find me agonizing about this on page 526, coming down on the side of 0.667, but not overwhelmingly convincingly. You could argue that a branch length should be set 0 when the branch is not there, and all the resulting values averaged, or you could argue that the average should only be taken over those trees for which that branch is present. One possible way to solve the problem is to take the consensus tree as if it were a user-defined tree, use your whole data set, and infer branch lengths on that tree. Daniel has already expressed his legitimate concerns in such a case as to whether it takes (for example) trifurcations as if they were real rather than an expression of our uncertainty. J.F. Joe Felsenstein j...@gs.washington.edu Department of Genome Sciences and Department of Biology, University of Washington, Box 355065, Seattle, WA 98195-5065 USA (from 1 October 2012 to 10 December 2012 on sabbatical leave at) Department of Statistics, University of California, Berkeley, 367 Evans Hall, Berkeley, CA 94710 [[alternative HTML version deleted]] ___ R-sig-phylo mailing list R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo -- Emmanuel Paradis IRD, Jakarta, Indonesia http://ape.mpl.ird.fr/ ___ R-sig-phylo mailing list R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
Re: [R-sig-phylo] Why no branch lengths on consensus trees?
Emmanuel Paradis wrote: If you are Bayesian, the trees sampled from an MCMC are here for estimation including of the branch lengths, so you use them to compute some sort of consensus topology as well as its branch lengths. So it makes sense that MrBayes can do a consensus tree with branch lengths. I endorse the rest of Emmanuel's advice but let me quibble with this one. The posterior on trees may not consist mostly of trees varying around a single consensus. If the posterior had, for example, two modes, each centered around a different tree, a single consensus tree might not be appropriate, and branch lengths computed by averaging lengths over the two modes might not be a good guide to what the trees in the posterior looked like. I don't know enough about MrBayes features to know whether they have some way around this. There is a similar issue with parsimony methods -- the set of most parsimonious trees may have a consensus, which may well not be a most parsimonious tree. People who see the consensus of most parsimonious trees may not realize that the particular tree they are looking at is not most parsimonious. J.F. Joe Felsenstein j...@gs.washington.edu Department of Genome Sciences and Department of Biology, University of Washington, Box 355065, Seattle, WA 98195-5065 USA (from 1 October 2012 to 10 December 2012 on sabbatical leave at) Department of Statistics, University of California, Berkeley, 367 Evans Hall, Berkeley, CA 94710 ___ R-sig-phylo mailing list R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
Re: [R-sig-phylo] Why no branch lengths on consensus trees?
Dear Scott, What should branch lengths on a consensus tree represent? They cannot be expected substitutions per residue. This would imply no evolution at points where uncertain branching patterns have been reduced to a multi-furcation - which is not what the multi-furcation is meant to imply. (Rather: there was evolution, but we aren't very certain about the branching pattern.) But, MrBayes does provide average lengths of some kind. Best wishes, Daniel On 21/11/2012 15:13, Scott Chamberlain myrmecocys...@gmail.com wrote: When making a consensus tree using ape::consensus the branch lengths are lost. Is there a way to not lose the branch lengths? Or to add them somehow to the consensus tree after making it. library(ape) cat(owls(((Strix_aluco:4.2,Asio_otus:4.1):4.1,Athene_noctua:7.3):6.3,Tyt o_alba:13.5);, file = ex1.tre, sep = \n) cat(owls(((Strix_aluco:1.2,Asio_otus:4.5):3.1,Athene_noctua:7.3):6.3,Tyt o_alba:13.5);, file = ex2.tre, sep = \n) cat(owls(((Strix_aluco:3.2,Asio_otus:4.7):8.1,Athene_noctua:7.3):6.3,Tyt o_alba:13.5);, file = ex3.tre, sep = \n) tree1 - read.tree(ex1.tre) tree2 - read.tree(ex2.tre) tree3 - read.tree(ex3.tre) trees - c(tree1, tree2, tree3) trees_con - consensus(trees) trees_con Phylogenetic tree with 4 tips and 3 internal nodes. Tip labels:[1] Strix_aluco Asio_otus Athene_noctua Tyto_alba Rooted; no branch lengths. Thanks, Scott Chamberlain [[alternative HTML version deleted]] ___ R-sig-phylo mailing list R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo -- Daniel Barker http://bio.st-andrews.ac.uk/staff/db60.htm The University of St Andrews is a charity registered in Scotland : No SC013532 ___ R-sig-phylo mailing list R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
Re: [R-sig-phylo] Why no branch lengths on consensus trees?
I have a function to create a consensus tree with branch lengths. You feed it a given topology (often a consensus topology, made with ape), then a list of trees, and tell it what you want the branch lengths to represent. It could be the proportion of input trees with that edge (good for summarizing bootstrap or Bayes proportions) or the mean, median, or sd of branch lengths for those trees that have that edge. Consensus branch lengths in units of proportion of matching trees has obvious utility. As Daniel says, the average branch lengths across a set of trees is more difficult to see a use case for, but you could imagine doing something like taking the ratogram output from r8s on a set of trees and summarizing the rate average and rate sd on a given, best, tree as two sets of branch lengths on that tree. I've put the function source at https://r-forge.r-project.org/scm/viewvc.php/*checkout*/pkg/R/consensusBrlen.R?revision=110root=omearalab. You can source the file for the function (consensusBrlen() ) and other functions it needs. It also uses phylobase. Note that this is alpha-quality code -- it's been checked a bit, but verify it's doing what you want. Here's an example of how to use it library(ape) library(phylobase) phy.a-rcoal(15) phy.b-phy.a phy.b$edge.length-phy.b$edge.length+runif(length(phy.b$edge.length), 0, 0.1) phy.c-rcoal(15) phy.list-list(phy.a, phy.b, phy.c) phy.consensus-consensusBrlen(phy.a, list(phy.a, phy.b, phy.c), type=mean_brlen) Best, Brian PS: Note that I am actively looking for grad students: info at http://www.brianomeara.info/lab . Guaranteed five years support, subject to decent performance. ___ Brian O'Meara Assistant Professor Dept. of Ecology Evolutionary Biology U. of Tennessee, Knoxville http://www.brianomeara.info Students wanted: Applications due Dec. 15, annually Postdoc collaborators wanted: Check NIMBioS' website Calendar: http://www.brianomeara.info/calendars/omeara On Wed, Nov 21, 2012 at 11:09 AM, Daniel Barker d...@st-andrews.ac.ukwrote: Dear Scott, What should branch lengths on a consensus tree represent? They cannot be expected substitutions per residue. This would imply no evolution at points where uncertain branching patterns have been reduced to a multi-furcation - which is not what the multi-furcation is meant to imply. (Rather: there was evolution, but we aren't very certain about the branching pattern.) But, MrBayes does provide average lengths of some kind. Best wishes, Daniel On 21/11/2012 15:13, Scott Chamberlain myrmecocys...@gmail.com wrote: When making a consensus tree using ape::consensus the branch lengths are lost. Is there a way to not lose the branch lengths? Or to add them somehow to the consensus tree after making it. library(ape) cat(owls(((Strix_aluco:4.2,Asio_otus:4.1):4.1,Athene_noctua:7.3):6.3,Tyt o_alba:13.5);, file = ex1.tre, sep = \n) cat(owls(((Strix_aluco:1.2,Asio_otus:4.5):3.1,Athene_noctua:7.3):6.3,Tyt o_alba:13.5);, file = ex2.tre, sep = \n) cat(owls(((Strix_aluco:3.2,Asio_otus:4.7):8.1,Athene_noctua:7.3):6.3,Tyt o_alba:13.5);, file = ex3.tre, sep = \n) tree1 - read.tree(ex1.tre) tree2 - read.tree(ex2.tre) tree3 - read.tree(ex3.tre) trees - c(tree1, tree2, tree3) trees_con - consensus(trees) trees_con Phylogenetic tree with 4 tips and 3 internal nodes. Tip labels:[1] Strix_aluco Asio_otus Athene_noctua Tyto_alba Rooted; no branch lengths. Thanks, Scott Chamberlain [[alternative HTML version deleted]] ___ R-sig-phylo mailing list R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo -- Daniel Barker http://bio.st-andrews.ac.uk/staff/db60.htm The University of St Andrews is a charity registered in Scotland : No SC013532 ___ R-sig-phylo mailing list R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo [[alternative HTML version deleted]] ___ R-sig-phylo mailing list R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
Re: [R-sig-phylo] Why no branch lengths on consensus trees?
Dear Brian, Awesome. Thanks for sharing the code Brian. I will give it a try. I see what you all mean now more precisely with the question of what does it really mean to have branch lengths on a consensus tree. Thanks, Scott On Wed, Nov 21, 2012 at 11:10 AM, Brian O'Meara bome...@utk.edu wrote: I have a function to create a consensus tree with branch lengths. You feed it a given topology (often a consensus topology, made with ape), then a list of trees, and tell it what you want the branch lengths to represent. It could be the proportion of input trees with that edge (good for summarizing bootstrap or Bayes proportions) or the mean, median, or sd of branch lengths for those trees that have that edge. Consensus branch lengths in units of proportion of matching trees has obvious utility. As Daniel says, the average branch lengths across a set of trees is more difficult to see a use case for, but you could imagine doing something like taking the ratogram output from r8s on a set of trees and summarizing the rate average and rate sd on a given, best, tree as two sets of branch lengths on that tree. I've put the function source at https://r-forge.r-project.org/scm/viewvc.php/*checkout*/pkg/R/consensusBrlen.R?revision=110root=omearalab. You can source the file for the function (consensusBrlen() ) and other functions it needs. It also uses phylobase. Note that this is alpha-quality code -- it's been checked a bit, but verify it's doing what you want. Here's an example of how to use it library(ape) library(phylobase) phy.a-rcoal(15) phy.b-phy.a phy.b$edge.length-phy.b$edge.length+runif(length(phy.b$edge.length), 0, 0.1) phy.c-rcoal(15) phy.list-list(phy.a, phy.b, phy.c) phy.consensus-consensusBrlen(phy.a, list(phy.a, phy.b, phy.c), type=mean_brlen) Best, Brian PS: Note that I am actively looking for grad students: info at http://www.brianomeara.info/lab . Guaranteed five years support, subject to decent performance. ___ Brian O'Meara Assistant Professor Dept. of Ecology Evolutionary Biology U. of Tennessee, Knoxville http://www.brianomeara.info Students wanted: Applications due Dec. 15, annually Postdoc collaborators wanted: Check NIMBioS' website Calendar: http://www.brianomeara.info/calendars/omeara On Wed, Nov 21, 2012 at 11:09 AM, Daniel Barker d...@st-andrews.ac.ukwrote: Dear Scott, What should branch lengths on a consensus tree represent? They cannot be expected substitutions per residue. This would imply no evolution at points where uncertain branching patterns have been reduced to a multi-furcation - which is not what the multi-furcation is meant to imply. (Rather: there was evolution, but we aren't very certain about the branching pattern.) But, MrBayes does provide average lengths of some kind. Best wishes, Daniel On 21/11/2012 15:13, Scott Chamberlain myrmecocys...@gmail.com wrote: When making a consensus tree using ape::consensus the branch lengths are lost. Is there a way to not lose the branch lengths? Or to add them somehow to the consensus tree after making it. library(ape) cat(owls(((Strix_aluco:4.2,Asio_otus:4.1):4.1,Athene_noctua:7.3):6.3,Tyt o_alba:13.5);, file = ex1.tre, sep = \n) cat(owls(((Strix_aluco:1.2,Asio_otus:4.5):3.1,Athene_noctua:7.3):6.3,Tyt o_alba:13.5);, file = ex2.tre, sep = \n) cat(owls(((Strix_aluco:3.2,Asio_otus:4.7):8.1,Athene_noctua:7.3):6.3,Tyt o_alba:13.5);, file = ex3.tre, sep = \n) tree1 - read.tree(ex1.tre) tree2 - read.tree(ex2.tre) tree3 - read.tree(ex3.tre) trees - c(tree1, tree2, tree3) trees_con - consensus(trees) trees_con Phylogenetic tree with 4 tips and 3 internal nodes. Tip labels:[1] Strix_aluco Asio_otus Athene_noctua Tyto_alba Rooted; no branch lengths. Thanks, Scott Chamberlain [[alternative HTML version deleted]] ___ R-sig-phylo mailing list R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo -- Daniel Barker http://bio.st-andrews.ac.uk/staff/db60.htm The University of St Andrews is a charity registered in Scotland : No SC013532 ___ R-sig-phylo mailing list R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo [[alternative HTML version deleted]] ___ R-sig-phylo mailing list R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
Re: [R-sig-phylo] Why no branch lengths on consensus trees?
Dear Joe, Thanks for your feedback on this question. I will go read those pages you mentioned. Scott On Wed, Nov 21, 2012 at 5:19 PM, Joe Felsenstein j...@gs.washington.eduwrote: Daniel Barker wrote: What should branch lengths on a consensus tree represent? Scott Chamberlain had written: When making a consensus tree using ape::consensus the branch lengths are lost. Is there a way to not lose the branch lengths? Or to add them somehow to the consensus tree after making it. The issue of what branch lengths ought to be on a consensus tree is not simple. If we have three rooted trees: ((A:1,B:1):1,C:2); ((A:1,B:1):1,C:2); (A:2,(B:1,C:1):1); the consensus tree should be the first tree, but what branch length should be used for (say) the branch ancestral to the AB clade? 1? 0.667? The minute you open this can of worms it becomes clear that the answer depends on what you want that number to convey and what interpretations your audience will tend to draw from the number. There is no obvious answer. So this is not a mere technical computing question. By the way, in my 2004 book, you will find me agonizing about this on page 526, coming down on the side of 0.667, but not overwhelmingly convincingly. You could argue that a branch length should be set 0 when the branch is not there, and all the resulting values averaged, or you could argue that the average should only be taken over those trees for which that branch is present. One possible way to solve the problem is to take the consensus tree as if it were a user-defined tree, use your whole data set, and infer branch lengths on that tree. Daniel has already expressed his legitimate concerns in such a case as to whether it takes (for example) trifurcations as if they were real rather than an expression of our uncertainty. J.F. Joe Felsenstein j...@gs.washington.edu Department of Genome Sciences and Department of Biology, University of Washington, Box 355065, Seattle, WA 98195-5065 USA (from 1 October 2012 to 10 December 2012 on sabbatical leave at) Department of Statistics, University of California, Berkeley, 367 Evans Hall, Berkeley, CA 94710 [[alternative HTML version deleted]] ___ R-sig-phylo mailing list R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
