Phylogenetic methods - computer practicals
Parsimony    Maximum likelihood    Bayesian inference    Species tree    home

Department of Botany, Charles University, Prague (2021-2024)

Parsimony analysis

Basic comments to the Nexus format are here.

Training dataset is here. This dataset includes a total of 1,127 AFLP scoring fragments for 81 individuals from 7 species of Veronica subsect. Pentasepalae (modified matrix from the one published in Padilla-García et al. 2018).
Samples are coded as following: SpeciesCode_ColectorNo_IndivNo_PCRcode. Example: Vara_NPG18_14_D11_13 is SpeciesCode: Vara, ColectorNo: NPG18, IndivNo: 14, PCR code: D11_13.

Download and install software:
PAUP* (4a169) - program for maximum parsimony inference and other methods)
FigTree - phylogenetic tree viewer

Other software recommended to install

• Notepad++
• Total Commander

Tasks (to work with Veronica Pentasepalae dataset)
after you are done submit the answers using this Google Form

1. Create a parsimony-based tree using PAUP (AFLP dataset)
PAUP has the advantage of being able to analyze data using several different optimality criteria; parsimony, likelihood, and distance. As you already know, each criterion has its strengths and limitations. To begin with, you will search for trees under the parsimony criterion (the default setting in PAUP).

• Open file and execute
  1. File -> Open… (select the NEXUS file „Veronica_example.nex“ and click Execute)
  2. Create a log file. It is a good idea to keep track of things that you are doing in PAUP by creating a log file. By default, PAUP will create a log file with the same name as the data file, but with a “.log” suffix on it. However, you can name it anything you want:
  - File -> Log Output To File
          - Click on Set... and Save as: parsimony.practice.log
          - if a file names “practice.log” already exists, you will be asked to Append, Cancel, or Replace
          - if that is the case, click Replace
          - now in the Log Output dialog, click OK
  

• Parsimony analysis
  1. Analysis -> Parsimony (Note: parsimony is the default setting and will probably already be selected).
  2. Analysis -> Heuristic Search
      Options:
          - General: Keep – optimal trees only, Set MaxTrees – 1000, Action if this limit is hit – Leave unchanged, and don't prompt
          - Starting trees – Get by stepwise addition
          - Stepwise addition: Addition sequence – random, # reps – 10, Show running status report
          - Branch swapping: swapping algorithm – TBR
  Question 1: PAUP provides two basic classes of methods for searching for optimal trees; exact, branch&bound and heuristic. Which are the main differences among these methods ? We selected the heuristic search. Why do you think this is the best option for using with this dataset?
  Question 2: Which algorithm are we using to obtain the trees?
  Question 3: Which algorithm are we using to reorganize the trees? Can you explain how does this algorith work?
  
  

• Rooting and visualizing trees
  1. Trees - > Root Trees…
      Rooting Options…
      - Define Outgroup…
      - double-click on samples of the outgroup (starting with Out). OK
          Out_Vori_MA3437_G04_
          Out_Vori_S218_H04_21
  Question 4: Why it is recommended to root the tree?
  2. Trees -> Print/View Trees…(it is not necessary to save these trees)
  

• Computing Consensus Tree
  1. Trees -> Compute Consensus…
      - Select Strict & Majority-rule (50%)
      - Include values in treefile… as node labels
      - Output to treefile - save „strict.majrule50.consensus.tre“
  2. Trees -> Print/View Consensus Tree(s)…
      - Print this Tree to PDF „strict.majrule50.consensus.tree.pdf“
  Open the pdf. Strict consensus tree appears in page 1 and majority-rule consensus tree in page 2.
  Question 5: Which differences do you observe between the two obtained consensus trees? What do the values observed in the majorityRule50 consensus tree are indicating?
  
  

• Bootstrap analysis
  1. Analysis -> Bootstrap/Jackknife Analysis
      - Resampling method – Bootstrap
      - Number of replicates 100 (normally a minimum of 1000 is recommended but we will use 100 to reduce the computational time)
      - Save trees for each replicate to file „bootstrap.replicates.tre“
  2. Trees -> Print/View Bootstrap Consensus to pdf „Bootstrap.consensus.tree.pdf“
  Question 6: What do bootstrap values mean?
  Question 7: Look and compare the „majorityRule50.consensus tree and the „bootstrap.consensus tree“. Why bootstrap values are different to the values obtained in the majority rule consensus tree?
  Question 8: According to these analyses, which species do you consider that are well-supported within Veronica Pentasepalae ? Why?

Tasks2 (to work with Amomum dataset, see Bayesian analysis)

1. Create a maximum parsimony tree of Amomum

• Open file etc.
  - File -> Open… (select a NEXUS file) or File -> Import Data … (allows to open FASTA, PHYLIP etc.)
  - File -> Execute (or CTRL + R)
  - File -> Log Output To Disk

• Parsimony analysis
  - Analysis -> Parsimony
  - Analysis -> Parsimony Settings…
      - Optimization – ACCTRAN or DELTRAN
      - Gaps – Missing Data
  - Analysis -> Heuristic Search
      - General: Keep – optimal trees only, Set MaxTrees – 1000, Leave unchanged
      - Starting trees – Get by stepwise addition
      - Stepwise addition: Addition sequence – random, # reps – 10
      - Branch swapping: swapping algorithm – TBR
  - Trees - > Root Trees…
      - Rooting Options… - Define Outgroup… - double-click on samples of the outgroup
        Siphonochilus_kirkii
  - Trees -> Print/View Trees…
  - Trees -> Compute Consensus… - Strict, Majority-rule (50%), Include values in treefile… as node labels, Output to treefile
  - Trees -> Print/View Consensus Tree(s)…

• Bootstrap
  - Analysis -> Bootstrap/Jackknife Analysis
      - Resampling method – Bootstrap
      - Number of replicates
  - Trees -> Print/View Bootstrap Consensus…

• Open the tree in FigTree

Thank you for participating in PAUP practicals...