Ancestral state reconstruction Molecular Phylogenetics exercise Ben Stver

Ancestral state reconstruction Molecular Phylogenetics – exercise Ben Stöver WS 2012/2013

Ancestal state reconstruction with Bayes. Traits 1. Ancestral state reconstruction 2 1 Ancestral state reconstruction • Reconstructing character states of acenstral nodes • Tree topology is given • Terminal character states are given • Top left charts: Female gametangia (green: ventral, red: marginal, blue: central) • Top right charts: Mycothallus (green: parenchymal, red: epidermal, blue: none) • Bottom charts: Scales (yellow: small, green: broad marginal, red: lateral on margin, blue: broad central) Ben Stöver 13. 02. 2012 Winter term 2012/2013 Fortgeschrittenenmodul Molecular Phylogenetics

Ancestal state reconstruction with Bayes. Traits 2. Mesquite 3 2 Mesquite • Modular software system • Ancestral state reconstruction with parsimony or likelihood • (No Bayesian methods available) • Many other features • Distributed under GPL Ben Stöver 13. 02. 2012 Winter term 2012/2013 Fortgeschrittenenmodul Molecular Phylogenetics

Ancestal state reconstruction with Bayes. Traits 3. Bayes. Traits 4 3. 1 Bayes. Traits – overview • Bayes. Traits is a command line based software (like PAUP* and Mr. Bayes) • Performs maximum likelihood or Bayesian reconstruction. Parts of the software: • Bayes. Multistate • Reconstructs characters with multiple (more than two) states • Bayes. Discrete • Analyze correlated evolution between pairs of discrete binary traits • Bayes. Continues • Analysis of the evolution of continuously varying traits (e. g. length of an organ) Ben Stöver 13. 02. 2012 Winter term 2012/2013 Fortgeschrittenenmodul Molecular Phylogenetics

Ancestal state reconstruction with Bayes. Traits 3. Bayes. Traits 5 3. 2 Starting Bayes. Traits • Bayes. Traits takes two command line parameters. • Bayes. Traits. exe <Tree. File> <Data. File> • The tree file may also contain a tree distribution (e. g. from Mr. Bayes) • The data file contains a table specifying the terminal character states. • In the start menu the part of the program to be used must be selected • The method is selectes afterwards Ben Stöver 13. 02. 2012 Winter term 2012/2013 Fortgeschrittenenmodul Molecular Phylogenetics

Ancestal state reconstruction with Bayes. Traits 3. Bayes. Traits 6 3. 3 Bayes. Traits commands • sample <n>: Sets the sample frequency. • iterations <n>: Specifies the number of generations for the MCMCalgorithm. • burnin <n>: Sets the generation count of the burin. • revjumphp: Specifies the distribution of prior probabilities (only for MCMC) • Add. Node: Defines an exact node to be reconstructed. (e. g. : Add. Node 01 51 52 50 49 47 48) • Add. MRCA: Defines a node to be reconstructed that is the most recent common ancestor of the specified taxa (but might also be the acestor of additional taxa. ) • fossil: Specifies an ancestral state already known. (e. g. : fossil Node 1 0 51 52 50 49 47 48 sets the state of Node 1=0) • info: Prints out information of the current settings. • run: Starts the analysis. Ben Stöver 13. 02. 2012 Winter term 2012/2013 Fortgeschrittenenmodul Molecular Phylogenetics

Ancestal state reconstruction with Bayes. Traits 3. Bayes. Traits 7 3. 4 Output of Bayes. Traits • Bayes. Traits generates a text file containing a table with the posterior probabilities of the samples (or likelihoods). • Each line describes one sample the MCMC algorithm took. • (For ML with a single tree, there will only be one line. ) • There is one column for each state of each character that was reconstructed. Ben Stöver 13. 02. 2012 Winter term 2012/2013 Fortgeschrittenenmodul Molecular Phylogenetics