La lecture à portée de main
Description
Sujets
Informations
Publié par | heinrich-heine-universitat_dusseldorf |
Publié le | 01 janvier 2003 |
Nombre de lectures | 7 |
Langue | English |
Extrait
Phylogenetic Trees from Large Datasets
Inaugural – Dissertation
zur
Erlangung des Doktorgrades der
Mathematisch–Naturwissenschaftlichen Fakult at
der Heinrich–Heine–Universitat Dusse ldorf
vorgelegt von
Heiko A. Schmidt
aus Bonn
Dusseldorf
2003Gedruckt mit der Genehmigung der Mathematisch-Naturwissenschaftlichen
Fakult at der Heinrich-Heine-Universitat Dusseldorf
Referent: Prof. Dr. Arndt von Haeseler
Korreferent: Prof. Dr. William Martin
Tag der mundlic hen Prufung: 28.07.2003The Dachshund
Indeed, it was made on the plan of a bench for length and lowness. It seemed
to be satis ed, but I thought the plan poor, and structurally weak, on account
of the distance between the forward supports and those abaft. With age the
dog’s back was likely to sag; and it seemed to me that it would have been a
stronger and more practicable dog if it had had some more legs.
Mark Twain (Following the Equator, 1895-1896)
Biologists must constantly keep in mind that what they see was not designed,
but rather evolved.
Francis Crick (What Mad Pursuit, 1988)iv
Preface
ThisthesisdealswithtopicsfromBioinformatics/Phyloinformatics. Ittouchesthe elds
algorithmic complexity, molecular phylogenetics, sequence evolution, as well as parallel
computing and, thus, is addressed to an interdisciplinary community. Although it is
hardly possible to write a completely self-contained thesis, I tried to provide some basic
information needed for members of either community to be able to more easily under-
standthetopicsdiscussed. Consequently,someinformationmightbeenclosedthatseem
trivial to, e.g., computer scientists but which biologists are possibly not familiar with
and vice versa.
When writing up my thesis I decided, as it is custom in scienti c presentations, to use
the scienti c ’we’, which I prefer over the personal pronoun ’I’.
Parts of this thesis have been published in the following articles:
1. H. A. Schmidt, K. Strimmer, M. Vingron, and A. von Haeseler (2002) TREE-
PUZZLE: Maximum likelihood phylogenetic analysis using quartets and parallel
computing. Bioinformatics, 18, 502-504.
2. H. A. Schmidt and A. von Haeseler (2002) Quartet Trees as a Tool to Reconstruct
Large Trees from Sequences. In K. Jajuga, A. Sokolowski, and H.-H. Bock (eds.),
Data Analysis, Classi cation, and Related Methods, Studies in Classi cation, Data
Analysis, and Knowledge Organization, pages 379-388, Springer, Heidelberg, New
York.
3. H. A. Schmidt and A. von Haeseler (2003) Maximum Likelihood Analysis Using
TREE-PUZZLE. In A. D. Baxevanis, D. B. Davison, R. D. M. Page, G. Stormo,
and L. Stein (eds.), Current Protocols in Bioinformatics, Unit 6.6, pages 6.6.1-
6.6.25, Wiley and Sons, New York.
4. H. A. Schmidt, E. Petzold, M. Vingron, and A. von Haeseler (2003) Molecular
Phylogenetics: Parallelized Parameter Estimation and Quartet Puzzling. J. Par-
allel Distrib. Comput., 63, in press.
The TREE-PUZZLE package including developments presented in this thesis is freely
available from http://www.tree-puzzle.de.Contents
1. Overview 1
1.1. Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2. Organization of this Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Molecular Phylogenetics: A General Introduction 5
2.1. Biological Sequences and Molecular Evolution . . . . . . . . . . . . . . . 5
2.1.1. Types of Biological Data . . . . . . . . . . . . . . . . . . . . . . . 5
2.1.2. Sequence Evolution . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.1.3. Multiple Sequence Alignment . . . . . . . . . . . . . . . . . . . . 7
2.1.4. Modeling Molecular Evolution . . . . . . . . . . . . . . . . . . . . 7
2.2. Phylogenetic Tree Reconstruction . . . . . . . . . . . . . . . . . . . . . . 10
2.2.1. Notation on Phylogenetic Trees . . . . . . . . . . . . . . . . . . . 10
2.2.2. Types of Phylogenetic Methods . . . . . . . . . . . . . . . . . . . 14
2.2.3. Maximum Likelihood on Phylogenetic Trees . . . . . . . . . . . . 14
2.2.4. Complexity of Phylogenetic Analysis . . . . . . . . . . . . . . . . 17
3. The Quartet Puzzling Algorithm 19
3.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.2. Methods from the TREE-PUZZLE Package . . . . . . . . . . . . . . . . 20
3.2.1. Likelihood Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.2.2. Quartet Puzzling . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2.2.1. ML Step . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2.2.2. Puzzling Step . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2.2.3. Consensus Step . . . . . . . . . . . . . . . . . . . . . . . 24
4. Improvement in Complexity of the Puzzling Step of QP 25
4.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.2. Complexity Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.3.y of the Original Puzzling Step . . . . . . . . . . . . . . . . . . 27
4.4. More E cient Puzzling Step Algorithms . . . . . . . . . . . . . . . . . . 29
4.4.1. Split-Based Puzzling Step Algorithm . . . . . . . . . . . . . . . . 29
4.4.2. Recursive Puzzling Step Algorithm . . . . . . . . . . . . . . . . . 31
4.4.2.1. Example . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.4.2.2. Complexity . . . . . . . . . . . . . . . . . . . . . . . . . 34
vvi Contents
4.4.3. Berry’s MRCA-Based Puzzling Step Algorithm . . . . . . . . . . . 36
4.5. Runtime Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.5.1. Datasets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.5.2. Benchmark Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.5.3. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.6. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
5. Parallelized Quartet Puzzling 43
5.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.2. Parameter Estimation for Evolutionary Models. . . . . . . . . . . . . . . 44
5.2.1. Algorithm: Estimating Model Parameters . . . . . . . . . . . . . 44
5.3. Datasets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5.4. Runtime Analysis of the Sequential TREE-PUZZLE . . . . . . . . . . . . 46
5.4.1. Runtime of the Parameter Estimation. . . . . . . . . . . . . . . . 46
5.4.2. Runtime of the Quartet Puzzling Algorithm . . . . . . . . . . . . 46
5.5. Parallelizing TREE-PUZZLE . . . . . . . . . . . . . . . . . . . . . . . . 47
5.5.1. Parallelizing the Parameter Estimation . . . . . . . . . . . . . . . 47
5.5.2. Parallelizing Quartet Puzzling . . . . . . . . . . . . . . . . . . . . 47
5.6. Scheduling Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
5.7. E ciency of Parallel TREE-PUZZLE . . . . . . . . . . . . . . . . . . . . 50
5.7.1. Benchmark Datasets and Setup . . . . . . . . . . . . . . . . . . . 50
5.7.2. Results for Parameter Estimation . . . . . . . . . . . . . . . . . . 51
5.7.3. ML Step and Puzzling Step . . . . . . . . . . . . . . . . . . . . . 51
5.7.4. Overall Scaling of Parallel TREE-PUZZLE . . . . . . . . . . . . . 51
5.8. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
6. Large ML Trees from Sequences Using Quartets 57
6.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
6.2. The Modi ed Quartet Puzzling Algorithm . . . . . . . . . . . . . . . . . 57
6.2.1. The Algorithm ModPUZZLE . . . . . . . . . . . . . . . . . . . 58
6.3. Computational Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
6.4. Some Practical Measures on the Method . . . . . . . . . . . . . . . . . . 60
6.5. Discussion and Possible Extensions . . . . . . . . . . . . . . . . . . . . . 61
7. Phylogenetic Trees from Multiple Genesets with Missing Data 63
7.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
7.2. Methods to Combine Datasets . . . . . . . . . . . . . . . . . . . . . . . . 64
7.2.1. Low Level Combination: Total Evidence . . . . . . . . . . . . . . 64
7.2.2. High Level Methods . . . . . . . . . . . . . . . . . . . . . . . . . 64
7.2.2.1. Combining Equal-Sized Datasets: Consensus Methods . 64
7.2.2.2. Com Overlapping Datasets: Supertreeds . . 66
7.3. Medium Level Combined Phylogenetic Analysis . . . . . . . . . . . . . . 68
7.3.1. Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
7.3.2. The Combined Quartet Method to Combine Genesets . . . . . . . 70Contents vii
7.3.2.1. Combining the ML Quartets . . . . . . . . . . . . . . . . 70
7.3.2.2. Computing the Overall Tree . . . . . . . . . . . . . . . . 71
7.3.2.3. Assessing Whether Genesets Can Be Combined . . . . . 71
7.3.2.4. Overlap-Guided Puzzling Step . . . . . . . . . . . . . . . 72
7.3.2.5. Relative Majority Consensus . . . . . . . . . . . . . . . 73
7.4. The Phylogeny of the Grasses . . . . . . . . . . . . . . . . . . . . . . . . 73
7.4.1. The Dataset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
7.4.2. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
7.4.2.1. Low Level Combination . . . . . . . . . . . . . . . . . . 74
7.4.2.2. High Level Combination . . . . . . . . . . . . . . . . . . 74
7.4.2.3. Medium Level Combina