The paper describing phyml is here, a brief interview with the authors is here. Maximum likelihood ratio test

Size: px

Start display at page:

Download "The paper describing phyml is here, a brief interview with the authors is here. Maximum likelihood ratio test"

Cynthia Beasley
6 years ago
Views:

1 equence alignment: CLUTALW MUCLE Removing ambiguous T-COFFEE FORBACK positions: Generation of pseudosamples: EQBOOT PROTDIT TREE-PUZZLE Calculating and PROTPAR PHYML evaluating phylogenies: NEIGHBOR FITCH H-TET in Comparing phylogenies: CONENE TREE-PUZZLE Comparing models: Maximum Likelihood Ratio Test Visualizing trees: ATV, njplot, treeview phyml PHYML - A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood An online interface is here ; there is a command line version that is described here (not as straight forward as in clustalw); a phylip like interface is automatically invoked, if you type phyml the manual is here. The paper describing phyml is here, a brief interview with the authors is here TreePuzzle ne PUZZLE TREE-PUZZLE is a very versatile maximum likelihood program that is particularly useful to analyze protein sequences. The program was developed by Korbian trimmer and Arnd von Haseler (then at the Univ. of Munich) and is maintained by von Haseler, Heiko A. chmidt, and Martin Vingron (contacts see ome possible pathways from sequence to tree, model and support values. TREE-PUZZLE allows fast and accurate estimation of ARV (through estimating the shape parameter alpha) for both nucleotide and amino acid sequences (see here for figures). It has a fast algorithm to calculate trees through quartet puzzling (calculating ml trees for quartets of species and building the multispecies tree from the quartets). The program provides confidence numbers (puzzle support values), which tend to be smaller than bootstrap values (i.e. provide a more conservative estimate), the program calculates branch lengths and likelihood for user defined trees, which is great if you want to compare different tree topologies, or different models using the maximum likelihood ratio test. Branches which are not significantly supported are collapsed. TREE-PUZZLE runs on "all" platforms TREE-PUZZLE reads PHYLIP format, and communicates with the user in a way similar to the PHYLIP programs. Maximum likelihood ratio test If you want to compare two models of evolution (this includes the tree) given a data set, you can utilize the so-called maximum likelihood ratio test. If L and L 2 are the likelihoods of the two models, d =2(logL -logl 2 ) approximately follows a Chi square distribution with n degrees of freedom. Usually n is the difference in model parameters. I.e., how many parameters are used to describe the substitution process and the tree. In particular n can be the difference in branches between two trees (one tree is more resolved than the other). In principle, this test can only be applied if on model is a more refined version of the other. In the particular case, when you compare two trees, one calculated without assuming a clock, the other assuming a clock, the degrees of freedom are the number of OTUs 2 (as all sequences end up in the present at the same level, their branches cannot be freely chosen). To calculate the probability you can use the CHIQUARE calculator for windows available from Paul Lewis. TREE-PUZZLE allows (cont) TREEPUZZLE calculates distance matrices using the ml specified model. These can be used in FITCH or Neighbor. PUZZLEBOOT automates this approach to do bootstrap analyses WARNING: this is a distance matrix analyses! The official script for PUZZLEBOOT is here you need to create a command file (puzzle.cmds), and puzzle needs to be envocable through the command puzzle. Your input file needs to be the renamed outfile from seqboot A slightly modified working version of puzzleboot_mod.sh is here, and here is an example for puzzle.cmds. Read the instructions before you run this! Maximum likelihood mapping is an excellent way to assess the phylogenetic information contained in a dataset. ML mapping can be used to calculate the support around one Puzzle is cool, don't leave home without ml mapping ml mapping ml mapping can asses the topology surrounding an individual branch : E.g.: If we want to know if Giardia lamblia forms the deepest branch within the known eukaryotes, we can use ML mapping to address this problem. To apply ml mapping we choose the "higher" eukaryotes as cluster a, another deep branching eukaryote (the one that competes against Giardia) as cluster b, Giardia as cluster c, and the outgroup as cluster d. For an example output see this sample ml-map. From: Olga Zhaxybayeva and J Peter Gogarten BMC Genomics 2002, 3: Figure 5. Likelihood-mapping analysis for two biological data sets. (Upper) The distribution patterns. (Lower) The occupancies (in percent) for the seven areas of attraction. (A) Cytochrome-b data from ref.. (B) Ribosomal DNA of major arthropod groups (5). From: Korbinian trimmer and Arndt von Haeseler Proc. Natl. Acad. ci. UA Vol., pp , June 7 An analysis of the carbamoyl phosphate synthetase domains with respect to the root of the tree of life is here.

ml mapping can asses the not necessarily treelike histories of genome Application of ML mapping to comparative Genome analyses see here for a comparison of different probability measures. Fig.

: Example and comparison of different measures see here for an approach that solves the problem of poor taxon sampling that is usually considered inherent with quartet analyses. Fig.

2 ml mapping can asses the not necessarily treelike histories of genome Application of ML mapping to comparative Genome analyses see here for a comparison of different probability measures. Fig. 3: outline of approach Fig. : Example and comparison of different measures see here for an approach that solves the problem of poor taxon sampling that is usually considered inherent with quartet analyses. Fig. 2: The principle of analyzing extended to obtain embedded quartets Example next slides: Cluster a: sequences outgroup (prokaryotes) Cluster b: 20 sequences other Eukaryotes Cluster c: sequences Plasmodium Cluster d: sequences Giardia (a,b)-(c,d) /\ / \ / 3 : 2 \ / : \ / \ (a,d)-(b,c) (a,c)-(b,d) Number of quartets in region : 68 (= 2.3%) Number of quartets in region 2: 2 (= 7.5%) Number of quartets in region 3: (= 68.2%) Occupancies of the seven areas, 2, 3,, 5, 6, 7: (a,b)-(c,d) /\ / \ / /\ \ / 6 \ / / 7 \ \ / \ / 3 : 5 : 2 \ / \ (a,d)-(b,c) (a,c)-(b,d) Number of quartets in region : 53 (= 8.%) Number of quartets in region 2: 5 (= 5.%) Number of quartets in region 3: 73 (= 6.8%) Number of quartets in region : 3 (=.%) Number of quartets in region 5: 0 (= 0.0%) Number of quartets in region 6: 26 (=.3%) Number of quartets in region 7: 0 (= 3.6%) TREE-PUZZLE PROBLEM/DRAWBACK The more species you add the lower the support for individual branches. While this is true for all algorithms, in TREE-PUZZLE this can lead to completely unresolved trees with only a few handful of sequences. Trees calculated via quartet puzzling are usually not completely resolved, and they do not correspond to the ML-tree: The determined multi-species tree is not the tree with the highest likelihood, rather it is the tree whose topology is supported through ml-quartets, and the lengths of the resolved branches is determined through maximum likelihood. puzzle example The best tree might not be the true tree. When can one conclude that a tree is a significantly worse explanation for the data compared to the best tree Estimate the probability that a dataset might have resulted from a given tree. Example: Kira s kangaroo data Usertrees - H test - go through outfile (PAR, PROTPAR and DNAPAR perform a similar test when confronted with multiple usertrees) Zhaxybayeva and Gogarten, BMC Genomics 2003 : 37 COMPARION OF DIFFERENT UPPORT MEAURE A: mapping of posterior probabilities according to trimmer and von Haeseler B: mapping of bootstrap support values C: mapping of bootstrap support values from extended ml-mapping versus More gene families group species according to environment than according to 6rRNA phylogeny In contrast, a themophilic archaeon has more genes grouping with the thermophilic bacteria bootstrap values from extended Reverend Thomas Bayes (702-76) Bayes Theorem Posterior Probability represents the degree to which we believe a given model accurately describes the situation given the available data and all of our prior information I Prior Probability describes the degree to which we believe the model accurately describes reality based on all of our prior information. Likelihood describes how well the model predicts the data P(data model, I) P(model data, I) = P(model, I) P(data,I) Normalizing constant Elliot ober s Gremlins Observation: Loud noise in the attic Hypothesis: gremlins in the attic playing bowling Likelihood = P(noise gremlins in the attic) P(gremlins in the attic noise) Alternative Approaches to Estimate Posterior Probabilities Bayesian Posterior Probability Mapping with MrBayes (Huelsenbeck and Ronquist, 200) Problem: trimmer s formula olution: Exploration of the tree space by sampling trees using a biased random walk (Implemented in MrBayes program) Trees with higher likelihoods will be sampled more often p i N i N total L i p i = L +L 2 +L 3 only considers 3 trees (those that maximize the likelihood for the three topologies),where N i - number of sampled trees of topology i, i=,2,3 N total total number of sampled trees (has to be large) 2

Illustration of a biased random walk Figure generated using MCRobot program (Paul Lewis, 200) Phylogenetic information present in pectral Decomposition of Phylogenetic Data Break information into

BIPARTITION OF A PHYLOGENETIC TREE Bipartition (or split) a division of a phylogenetic tree into two parts that are connected by a single branch.

transferred gene (virulence factor homologs (mvin)) Lento -plot of supported bipartitions (out of 50 possible) 3 gammaproteobacterial (258 putative orthologs): E.

Prochlorococcus marinus (3 strains) Marine ynechococcus Thermosynechococcus elongatus Gloeobacter Nostoc punctioforme Number of There are 3,7,30,575 possible unrooted tree topologies for 3 only 258

3 Illustration of a biased random walk Figure generated using MCRobot program (Paul Lewis, 200) Phylogenetic information present in pectral Decomposition of Phylogenetic Data Break information into small quanta of information (bipartitions or embedded quartets) Analyze spectra to detect transferred genes and plurality consensus. BIPARTITION OF A PHYLOGENETIC TREE Bipartition (or split) a division of a phylogenetic tree into two parts that are connected by a single branch. It divides a dataset into two groups, but it does not consider the relationships within each of the two groups. Yellow vs Rest * * *... * * 5 compatible to illustrated bipartition * * *..... Orange vs Rest.. *.... * incompatible to illustrated bipartition Lento -plot of 3 supported bipartitions (out of 082 possible) Consensus clusters of eight significantly supported bipartitions Phylogeny of putatively transferred gene (virulence factor homologs (mvin)) Lento -plot of supported bipartitions (out of 50 possible) 3 gammaproteobacterial (258 putative orthologs): E.coli Buchnera Haemophilus Pasteurella almonella Yersinia pestis (2 strains) Vibrio Xanthomonas (2 sp.) Pseudomonas Wigglesworthia 0 cyanobacteria: Anabaena Trichodesmium ynechocystis sp. Prochlorococcus marinus (3 strains) Marine ynechococcus Thermosynechococcus elongatus Gloeobacter Nostoc punctioforme Number of There are 3,7,30,575 possible unrooted tree topologies for 3 only 258 genes analyzed Based on 678 sets of orthologous genes Zhaxybayeva, Lapierre and Gogarten, Trends in Genetics, 200, 20(5): Consensus clusters corresponding to three significantly supported bipartitions Zhaxybayeva, Lapierre and Gogarten, Trends in Genetics, 200, 20(5): The phylogeny of ribulose bisphosphate carboxylase large subunit Example of bipartition analysis for five of photosynthetic bacteria (88 gene families) R Ct Ca Ct total 0 bipartitions R: Rhodobacter capsulatus, H: Heliobacillus mobilis, : ynechocystis sp., Ct: Chlorobium tepidum, Ca: Chloroflexus aurantiacus H R Ca H Plurality Chl. Biosynth. Bipartitions supported by genes from chlorophyll biosynthesis pathway Zhaxybayeva, Hamel, Raymond, and Gogarten, Genome Biology 200, 5: R20 Phylogenetic Analyses of Genes from chlorophyll biosynthesis pathway (extended ) Xiong et al. cience, :72-30 R: Rhodobacter capsulatus, H: Heliobacillus mobilis, : ynechocystis sp., Ct: Chlorobium tepidum, Ca: Chloroflexus aurantiacus Zhaxybayeva, Hamel, Raymond, and Gogarten, Genome Biology 200, 5: R20 3

PROBLEM WITH BIPARTITION No easy way to incorporate gene families that are not represented in all.

A single misplaced sequence can destroy all bipartitions. Bootstrap support values for embedded quartets Quartet spectral analyses of iterates over three loops: Repeat for all bootstrap samples.

+ : tree calculated from one pseudosample generated by bootstraping from an alignment of one gene family present in : embedded quartet for,,, and 0.

2006, Genome Research, in press Iterating over Bootstrap amples This gene family for the quartet of species A, B, C, D upports the Topology ((A, D), B, C) with 70% bootstrap support Bootstrap support

analyses ummary of phylogenetic information for one genome quartet for all gene families Total number of gene families containing the species quartet Quartet pectrum of cyanobacterial 330 possible

0 0 0 Number of gene families supporting the same topology as the plurality (colored according to bootstrap support level) Number of gene families supporting one of the two alternative quartet

4 PROBLEM WITH BIPARTITION No easy way to incorporate gene families that are not represented in all. The more sequences are added, the shorter the internal branches become, and the lower is the bootstrap support for the individual bipartitions. A single misplaced sequence can destroy all bipartitions. Bootstrap support values for embedded quartets Quartet spectral analyses of iterates over three loops: Repeat for all bootstrap samples. Repeat for all possible embedded quartets. Repeat for all gene families. + : tree calculated from one pseudosample generated by bootstraping from an alignment of one gene family present in : embedded quartet for,,, and 0. This bootstrap sample supports the topology ((,),,0) Zhaxybayeva et al. 2006, Genome Research, in press Iterating over Bootstrap amples This gene family for the quartet of species A, B, C, D upports the Topology ((A, D), B, C) with 70% bootstrap support Bootstrap support values for embedded quartets + : tree calculated from one pseudosample generated by bootstraping from an alignment of one gene family present in Illustration of one component of a quartet spectral analyses ummary of phylogenetic information for one genome quartet for all gene families Total number of gene families containing the species quartet Quartet pectrum of cyanobacterial 330 possible quartets Quartet spectral analyses of iterates over three loops: Repeat for all bootstrap samples. Repeat for all possible embedded quartets. Repeat for all gene families. : embedded quartet for,,, and 0. This bootstrap sample supports the topology ((,),,0) Number of gene families supporting the same topology as the plurality (colored according to bootstrap support level) Number of gene families supporting one of the two alternative quartet topologies 28 from relaxed core (core + with one or two taxa missing) 685 show conflicts with plurality Number of quartets PLURALITY IGNAL Gloeobacter marine ynechococcus 3Prochlorococcus N 2Prochlorococcus A Prochlorococcus Nostoc m 3P 2P P G Th Conflicts with plurality signal are observed in sets of orthologs across all functional categories, including genes involved in translation and transcription Genes with orthologs outside the cyanobacterial phylum: Distribution among Functional Categories (using COG db, release of March 2003) Anabaena Cyanobacteria do form a coherent group, but conflict with plurality (2) Cyanobacteria do not form a coherent group (60) 700 phylogenetically useful extended Trichodesmium Tr Crocosphaera C ynechocystis Thermosynechococcus 62/28 55%

Example of interphylum transfer: threonyl trna synthetase pecies evolution versus plurality consensus In case of the marine

The Coral of Life (Darwin) This is probably due to frequent gene transfer mediated by phages e.g.: These conflicting observations are not limited to prokaryotes.

behavior as belonging to the same species genetically more similar to a sister species (Grant et al.

588-5). Coalescence the process of tracing lineages backwards in time to their common ancestors.

t/2 (Kingman, 82) Illustration is from J.

one speciation event per generation REULT: One horizontal transfer event once in Most recent common ancestors are different for

erations (I.e., speciation events) molecular phylogenies RED: organismal lineages (no HGT) Different coalescence times BLUE: molecular

(2000) Nat Genet 26, 358-6 Albrecht Dürer, The Fall of Man, 50 Adam and Eve never met Mitochondrial Eve Lived 66,000-2,000 years ago

EXTANT LINEAGE FOR THE IMULATION OF 50 LINEAGE log (number of surviving lineages) Lineages Through Time Plot 0 simulations of

5 Example of interphylum transfer: threonyl trna synthetase pecies evolution versus plurality consensus In case of the marine ynecchococcus and Prochlorococcus spp. the plurality consensus is unlikely to reflect organismal history. The Coral of Life (Darwin) This is probably due to frequent gene transfer mediated by phages e.g.: These conflicting observations are not limited to prokaryotes. In incipient species of Darwin s finches frequent introgression can make some individuals characterized by morphology and mating behavior as belonging to the same species genetically more similar to a sister species (Grant et al. 200 Convergent evolution of Darwin's finches caused by introgressive hybridization and selection Evolution Int J Org Evolution 58, 588-5). Coalescence the process of tracing lineages backwards in time to their common ancestors. Every two extant lineages coalesce to their most recent common ancestor. Eventually, all lineages coalesce to the cenancestor. t/2 (Kingman, 82) Illustration is from J. Felsenstein, Inferring Phylogenies, inauer, 2003 Coalescence of ORGANIMAL and MOLECULAR Lineages Time 20 lineages One extinction and one speciation event per generation REULT: One horizontal transfer event once in Most recent common ancestors are different for organismal and 5 generations (I.e., speciation events) molecular phylogenies RED: organismal lineages (no HGT) Different coalescence times BLUE: molecular lineages (with HGT) GRAY: extinct lineages Long coalescence time for the last two lineages Y chromosome Adam Lived approximately 50,000 years ago Thomson, R. et al. (2000) Proc Natl Acad ci U A 7, Underhill, P.A. et al. (2000) Nat Genet 26, Albrecht Dürer, The Fall of Man, 50 Adam and Eve never met Mitochondrial Eve Lived 66,000-2,000 years ago Cann, R.L. et al. (87) Nature 325, 3-6 Vigilant, L. et al. () cience 253, The same is true for ancestral rrnas, EF, ATPases! EXTANT LINEAGE FOR THE IMULATION OF 50 LINEAGE log (number of surviving lineages) Lineages Through Time Plot 0 simulations of organismal evolution assuming a constant number of species (200) throughout the simulation; speciation and extinction per time step. (green O) 25 gene histories simulated for each organismal history assuming HGT per 0 speciation events (red x) green: organismal lineages ; red: molecular lineages (with gene transfer) Bacterial 6rRNA based phylogeny (from P. D. chloss and J. Handelsman, Microbiology and Molecular Biology Reviews, December 200.) The deviation from the long branches at the base pattern could be due to under sampling an actual radiation due to an invention that was not transferred following a mass extinction 5

New Tools for Visualizing Genome Evolution

New Tools for Visualizing Genome Evolution Lutz Hamel Dept. of Computer Science and Statistics University of Rhode Island J. Peter Gogarten Dept. of Molecular and Cell Biology University of Connecticut