Construc)ng the Tree of Life: Divide-and-Conquer! Tandy Warnow University of Illinois at Urbana-Champaign

Transcription

1 Construc)ng the Tree of Life: Divide-and-Conquer! Tandy Warnow University of Illinois at Urbana-Champaign

2 Phylogeny (evolutionary tree) Orangutan Gorilla Chimpanzee Human From the Tree of the Life Website, University of Arizona

3 Phylogenies and Applications Basic Biology: How did life evolve? Applica)ons of phylogenies to: protein structure and func)on popula)on gene)cs human migra)ons metagenomics Figure from hhps://en.wikipedia.org/wiki/common_descent

4 phylogenomics Orangutan Chimpanzee gene 1 gene 2 gene 999 gene 1000 Gorilla Human ACTGCACACCG ACTGC-CCCCG AATGC-CCCCG -CTGCACACGG CTGAGCATCG CTGAGC-TCG ATGAGC-TC- CTGA-CAC-G AGCAGCATCGTG AGCAGC-TCGTG AGCAGC-TC-TG C-TA-CACGGTG CAGGCACGCACGAA AGC-CACGC-CATA ATGGCACGC-C-TA AGCTAC-CACGGAT gene here refers to a portion of the genome (not a functional gene) I ll use the term gene to refer to c-genes : recombination-free orthologous stretches of the genome 2

5 DNA Sequence Evolution AAGGCCT AAGACTT TGGACTT -3 mil yrs -2 mil yrs AGGGCAT TAGCCCT AGCACTT -1 mil yrs AGGGCAT TAGCCCA TAGACTT AGCACAA AGCGCTT today

6 Phylogenetic Tree Estimation U V W X Y AGGGCAT TAGCCCA TAGACTT TGCACAA TGCGCTT U X Y V W

7 However U V W X Y AGGGCATGA AGAT TAGACTT TGCACAA TGCGCTT U X Y V W

8 Indels (insertions and deletions) Deletion Mutation ACGGTGCAGTTACCA ACCAGTCACCA

9 Deletion Substitution ACGGTGCAGTTACCA Insertion ACCAGTCACCTA ACGGTGCAGTTACC-A AC----CAGTCACCTA The true mul*ple alignment Reflects historical substitution, insertion, and deletion events Defined using transitive closure of pairwise alignments computed on edges of the true tree

10 Phylogenetic Tree Estimation S1 = AGGCTATCACCTGACCTCCA S2 = TAGCTATCACGACCGC S3 = TAGCTGACCGC S4 = TCACGACCGACA

11 Input: unaligned sequences S1 = AGGCTATCACCTGACCTCCA S2 = TAGCTATCACGACCGC S3 = TAGCTGACCGC S4 = TCACGACCGACA

12 Phase 1: Alignment S1 = AGGCTATCACCTGACCTCCA S2 = TAGCTATCACGACCGC S3 = TAGCTGACCGC S4 = TCACGACCGACA S1 = -AGGCTATCACCTGACCTCCA S2 = TAG-CTATCAC--GACCGC-- S3 = TAG-CT GACCGC-- S4 = TCAC--GACCGACA

13 Phase 2: Construct tree S1 = AGGCTATCACCTGACCTCCA S2 = TAGCTATCACGACCGC S3 = TAGCTGACCGC S4 = TCACGACCGACA S1 = -AGGCTATCACCTGACCTCCA S2 = TAG-CTATCAC--GACCGC-- S3 = TAG-CT GACCGC-- S4 = TCAC--GACCGACA S1 S2 S4 S3

14 Two-phase es)ma)on Alignment methods Clustal POY (and POY*) Probcons (and Probtree) Probalign MAFFT Muscle Di-align T-Coffee Prank (PNAS 2005, Science 2008) Opal (ISMB and Bioinf. 2007) FSA (PLoS Comp. Bio. 2009) Infernal (Bioinf. 2009) Etc. Phylogeny methods Bayesian MCMC Maximum parsimony Maximum likelihood Neighbor joining FastME UPGMA Quartet puzzling Etc.

15 Two-phase es)ma)on Alignment methods Clustal POY (and POY*) Probcons (and Probtree) Probalign MAFFT Muscle Di-align T-Coffee Prank (PNAS 2005, Science 2008) Opal (ISMB and Bioinf. 2007) FSA (PLoS Comp. Bio. 2009) Infernal (Bioinf. 2009) Etc. Phylogeny methods Bayesian MCMC Maximum parsimony Maximum likelihood Neighbor joining FastME UPGMA Quartet puzzling Etc. RAxML: heuris>c for large-scale ML op>miza>on

16 Quan)fying Error FN FN: false negative (missing edge) FP: false positive (incorrect edge) 50% error rate FP

17 1000-taxon models, ordered by difficulty (Liu et al., Science 19 June 2009)

18 Multiple Sequence Alignment (MSA): a scientific grand challenge 1 S1 = AGGCTATCACCTGACCTCCA S2 = TAGCTATCACGACCGC S3 = TAGCTGACCGC Sn = TCACGACCGACA S1 = -AGGCTATCACCTGACCTCCA S2 = TAG-CTATCAC--GACCGC-- S3 = TAG-CT GACCGC-- Sn = TCAC--GACCGACA Novel techniques needed for scalability and accuracy NP-hard problems and large datasets Current methods do not provide good accuracy Few methods can analyze even moderately large datasets Many important applications besides phylogenetic estimation 1 Frontiers in Massive Data Analysis, National Academies Press, 2013

19 1KP: Thousand Transcriptome Project G. Ka-Shu Wong U Alberta J. Leebens-Mack U Georgia N. Wickett Northwestern N. Matasci iplant T. Warnow, S. Mirarab, N. Nguyen UT-Austin UT-Austin UT-Austin l First publica)on: WickeH, Mirarab, et al., PNAS, 2014 l Used SATé (Liu et al., Science 2009 and Syst Biol 2012) to compute mul)ple sequence alignments and trees l Used ASTRAL (Mirarab et al., Bioinf 2014 and 2015) to compute the species tree Upcoming Challenge: Mul)ple sequence alignment and gene tree es)ma)on on 100,000 sequences

20 Computa*onal Phylogene*cs (2005) Current methods can use months to estimate trees on 1000 DNA sequences Our objective: More accurate trees and alignments on 500,000 sequences in under a week Courtesy of the Tree of Life web project, tolweb.org

21 Computa*onal Phylogene*cs (2015) : Distance-based phylogenetic tree estimation from polynomial length sequences 2012: Computing accurate trees (almost) without multiple sequence alignments : Co-estimation of multiple sequence alignments and gene trees, now on 1,000,000 sequences in under two weeks : Species tree estimation from whole genomes in the presence of massive gene tree heterogeneity Courtesy of the Tree of Life web project, tolweb.org

22 Computa*onal Phylogene*cs (2015) : Distance-based phylogenetic tree estimation from polynomial length sequences 2012: Computing accurate trees (almost) without multiple sequence alignments : Co-estimation of multiple sequence alignments and gene trees, now on 1,000,000 sequences in under two weeks : Species tree estimation from whole genomes in the presence of massive gene tree heterogeneity Courtesy of the Tree of Life web project, tolweb.org

23 Key technique: Divide-and-conquer! In general, small datasets with not too much heterogeneity are easy to analyze with good accuracy.

24 Divide-and-Conquer Divide-and-conquer is a basic algorithmic trick for solving problems! Three steps: divide a dataset into two or more sets, solve the problem on each set, and combine solu)ons.

25 Sor)ng Objec)ve: sort this list of integers from smallest to largest. 10, 3, 54, 23, 75, 5, 1, 25 should become 1, 3, 5, 10, 23, 25, 54, 75

26 MergeSort Step 1: Divide into two sublists Step 2: Recursively sort each sublist Step 3: Merge the two sorted sublists

27 Step 1: break into two lists X: Y:

28 Step 2: sort the two lists X: Y:

29 Step 3: merge the sorted lists X: Y: Result:

30 Merging (cont.) X: Y: Result: 1

31 Merging (cont.) X: Y: Result: 1 3

32 Merging (cont.) X: Y: Result: 1 3 5

33 Merging (cont.) X: Y: Result:

34 Merging (cont.) X: Y: Result:

35 Merging (cont.) X: Y: Result:

36 Merging (cont.) X: Y: 75 Result:

37 Merging (cont.) X: Y: Result:

38 Multiple Sequence Alignment (MSA): a scientific grand challenge 1 S1 = AGGCTATCACCTGACCTCCA S2 = TAGCTATCACGACCGC S3 = TAGCTGACCGC Sn = TCACGACCGACA S1 = -AGGCTATCACCTGACCTCCA S2 = TAG-CTATCAC--GACCGC-- S3 = TAG-CT GACCGC-- Sn = TCAC--GACCGACA Novel techniques needed for scalability and accuracy NP-hard problems and large datasets Current methods do not provide good accuracy Few methods can analyze even moderately large datasets Many important applications besides phylogenetic estimation 1 Frontiers in Massive Data Analysis, National Academies Press, 2013

39 SATé and PASTA Input: set of unaligned sequences Output: mul)ple sequence alignment and phylogene)c tree SATé: Liu et al., Science 2009 (up to 10,000 sequences) and Systema)c Biology 2012 (up to 50,000 sequences) PASTA: Mirarab et al., J. Comp Biol 2015 (up to 1,000,000 sequences)

40 1000-taxon models, ordered by difficulty (Liu et al., Science 19 June 2009)

41 Re-aligning on a tree A B C D Decompose dataset A C B D Align subproblems A C B D Estimate ML tree on merged alignment ABCD Merge subalignments

42 SATé and PASTA Algorithms Obtain initial alignment and estimated ML tree Tree Use tree to compute new alignment

43 SATé and PASTA Algorithms Obtain initial alignment and estimated ML tree Tree Use tree to compute new alignment Alignment

44 SATé and PASTA Algorithms Obtain initial alignment and estimated ML tree Estimate ML tree on new alignment Tree Use tree to compute new alignment Alignment

45 SATé and PASTA Algorithms Obtain initial alignment and estimated ML tree Estimate ML tree on new alignment Tree Use tree to compute new alignment Alignment Repeat un)l termina)on condi)on, and return the alignment/tree pair with the best ML score

46 SATé: 24-hour co-es*ma*on of highly accurate alignments and trees on 1000 sequences 1000-taxon models, ordered by difficulty (Liu et al., Science 19 June 2009) 24-hour SATé analysis, on desktop machines (Similar improvements for biological datasets)

47 SATé-2: even more accurate! (Liu et al., Syst Biol 61(1):90-106, 2012)

48 PASTA: even more accurate, and can scale to 1,000,000 sequences RNASim 0.20 Tree Error (FN Rate) Clustal Omega Muscle Mafft Starting Tree SATe2 PASTA Reference Alignment Simulated RNASim datasets from 10K to 200K taxa Limited to 24 hours using 12 CPUs Not all methods could run (missing bars could not finish) PASTA, Mirarab et al., J Comp Biol 22(5): (2015)

49 Avian Phylogenomics Project E Jarvis, HHMI MTP Gilbert, Copenhagen G Zhang, BGI T. Warnow UT-Aus)n S. Mirarab Md. S. Bayzid, UT-Aus)n UT-Aus)n First analysis (Jarvis, Mirarab, et al., Science 2014): Plus many many other people Approx. 50 species, 14,000 loci Used SATé for gene sequence alignment and tree es)ma)on Next analysis will have more species, and will use PASTA

50 1KP: Thousand Transcriptome Project G. Ka-Shu Wong U Alberta J. Leebens-Mack U Georgia N. Wickett Northwestern N. Matasci iplant T. Warnow, S. Mirarab, N. Nguyen UT-Austin UT-Austin UT-Austin First analysis (WickeH, Mirarab, et al., PNAS, 2014) About 100 species and 800 loci Used SATé Next analysis will be much larger and more difficult: Mul)ple sequence alignment and gene tree es)ma)on on 100,000 sequences, many datasets highly fragmentary Will use PASTA and UPP (Nguyen et al., Genome Biology 2015)

51 Computa*onal Phylogene*cs (2015) : Distance-based phylogenetic tree estimation from polynomial length sequences 2012: Computing accurate trees (almost) without multiple sequence alignments : Co-estimation of multiple sequence alignments and gene trees, now on 1,000,000 sequences in under two weeks : Species tree estimation from whole genomes in the presence of massive gene tree heterogeneity Courtesy of the Tree of Life web project, tolweb.org

52 Boosters, or Meta-Methods Meta-methods use divide-and-conquer and iteration (or other techniques) to boost the performance of base methods (phylogeny reconstruction, alignment estimation, etc) Base method M Meta-method M*

53 Main Points Innova)ve algorithm design can improve accuracy as well as reduce running *me. Divide-and-conquer is a key algorithmic technique that has drama)cally changed the toolkit for biologists!

54 Acknowledgments Funding: Guggenheim Founda)on, Packard Founda)on, NSF, Microsos Research New England, David Bruton Jr. Centennial Professorship, Grainger Founda)on, and TACC (Texas Advanced Compu)ng Center)

55 Avian Phylogenomics Project E Jarvis, HHMI MTP Gilbert, Copenhagen G Zhang, BGI T. Warnow UT-Aus)n S. Mirarab Md. S. Bayzid, UT-Aus)n UT-Aus)n Jarvis, Mirarab, et al., Science 2014 Major challenge: Plus many many other people Massive gene tree heterogeneity consistent with incomplete lineage sor)ng Very poor resolu)on in the 14,000 gene trees Standard coalescent-based species tree es)ma)on methods had poor accuracy Solu*on: New technique to improve coalescent-based species tree (sta*s*cal binning, Mirarab et al., Science 2014)