PAUP

PAUP (Phylogentic analysis using parsimony…and other methods) is one of the most highly used phylogenetic software programs. The new version does distance, parsimony and maximum likelihood methods. This sheet will give you an a rough idea of how to use the program, but a little knowledge can be dangerous. If you don’t understand what you are doing, ask someone in the lab who does.

We often use PAUP in conjunction with MODELTEST, a program that evaluates the best molecular model of evolution for your data

  1. Open your file in PAUP and select Edit.
  2. Next, OPEN then file PAUP4 Modelblock, and hit EDIT.
  3. Highlight and copy the text of this file and paste THE ENTIRE TEXT at the very bottom of your file file.
  4. Save the file under a new file name.
  5. Under FILE, select OPEN, select this new file, and EXECUTE. A screen should come up with a bunch of text that will change frequently. It may take 5-10 minutes (or longer) to complete. You will know when it’s done because it will say “processing of file completed”.
  6. When run is complete, a file called MODEL.SCORES will be created in the same folder that the original file was in.
  7. Close PAUP
  8. Open MODELTEST
  9. Click INPUT FROM FILE and specify the MODEL.SCORES file that was just created.
  10. Click OUTPUT TO FILE and specify a name and location for the file.
  11. The program will run in a fraction of a second and the output file will be created.
  12. In PAUP, open your file and the output file that you created from ModelTest. The later will have a block of likelihood scores and a list of hierarchical of likelihood ratio test results. At the base of this file is “Model Selected is:” This is your best model and is followed by all of the model parameters that best fit your data.
  13. COPY the model block from “[!Likelihood settings…” to “end;” (be sure to include [ ] and ;)
  14. PASTE the modelblock to the very bottom of your original PAUP file (the one without the modelblock), then save.
  15. EXECUTE your file in PAUP.
  16. In DATA, go to DEFINE OUTGROUP. Select the sequences from your outgroup. Often you can get outgroup senquences from Genbank.
  17. In OPTIONS, go to ROOTING
  18. Select MAKE INGROUP MONOPHYLETIC, hit OK

DISTANCE METHODS

  1. Under ANALYSIS, go to DISTANCE
  2. Next under ANALYSIS, go to DISTANCE SETTINGS
  3. Under DNA/RNA distances, select the distance metric that you’d like to use.
  4. In ANALYSIS, go to NEIGHBORJOINING/UPGMA. Select Neighborjoining and hit OK.
  5. Under TREES, go to TREE SCORES. Select Distance and write down the tree score.
  6. In TREES, go to PRINT NJ TREE
  7. In PLOT TYPE, Select PHYLOGRAM
  8. Hit PREVIEW
  9. Save tree as PICT file. Also, print the tree.

PARSIMONY ANALYSIS

  1. Next, in ANALYSIS, change to PARSIMONY
  2. In ANALYSIS go to HEURISTIC SEARCH
  3. Press SEARCH
  4. If there is more than one most parsimonious tree, in TREES, go to COMPUTE CONSENSUS.
  5. Select STRICT
  6. In TREES select PRINT CONSENSUS
  7. In PLOT TYPE, Select PHYLOGRAM
  8. Hit PREVIEW
  9. Save tree as PICT file. Also, print the tree.

LIKELIHOOD ANALYSIS

  1. In ANALYSIS, change optimality criterion to LIKELIHOOD.
  2. In ANALYSIS go to LIKELIHOOD SETTINGS.
  3. In ANALYSIS, go to HEURISTIC SEARCH.
  4. The model parameters of your data should be entered automatically if you put the PAUP block in appropriately from ModelTest.
  5. Hit SEARCH. This may take a very long time, be prepared to wait.
  6. Again, print the tree and save it as a pict file, as done before. Like parsimony there may be multiple trees and you may need to compute a consensus.

BOOTSTRAPPING

  1. EXECUTE your data file
  2. Define your outgroup, as done before.
  3. Set optimality criterion to PARSIMONY
  4. Under ANALYSIS, go to BOOTSTRAP/JACKKNIFE
  5. Select BOOTSTRAP, and set it for 100 Replicates. Retain all groups will Frequencies greater than 60%.
  6. Hit Continue and Search
  7. When finished, in TREES, go to PRINT BOOTSTRAP CONSENSUS
  8. Select Rectangular Cladogram. Select Show Group Frequencies
  9. Print the bootstrap consensus tree

KISHINO HASEGAWA TEST

  1. In MacClade, open your data file.
  2. In DISPLAY, go to GO TO TREE WINDOW
  3. Select DISPLAY RANDOM BUSH
  4. Dragging individual branches, create two groups of sequences for each population (hypothesis: populations are monophyletic groups)
  5. In the TOOLS menu, find the collapse clade tool and collapse each population clade into monophyletic polytomies, one for each population
  6. In TREE menu, go to STORE TREE and save as MONOPHYLY CONTRAINT
  7. Save the MacClade file
  8. Under TREE save tree file as Monophyly constraint. Close file
  9. Open Allsilversides in PAUP and EXECUTE
  10. Define your outgroup
  11. Set optimality criterion to PARSIMONY
  12. In ANALYSIS, go to LOAD CONSTRAINT
  13. Open MONOPHYLY CONTRAINT
  14. In ANAYLSIS, go to HEURISTIC SEARCH
  15. In GENERAL SEARCH OPTIONS, click ENFORCE TOPOLOGICAL CONSTRAINTS, keeping trees that ARE COMPATIBLE with constraints
  16. Hit SEARCH
  17. Save tree to file HK TEST
  18. In ANAYLSIS, go to HEURISTIC SEARCH
  19. In GENERAL SEARCH OPTIONS, unclick ENFORCE TOPOLOGICAL CONSTRAINTS, so that the heuristic search will no longer enforce the constraint tree.
  20. Hit SEARCH
  21. Save trees to the same file HK TEST. When asked, say APPEND FILE. You should now have three trees in the HK TEST treefile
  22. Open the HK TEST treefile
  23. Rather than adding all the trees together in one tree block, APPENDING the tree file merges two independent files together. You must now edit the file to create a single tree block. First, label the tree(s) that you got with the monophyly constraint so that you know what is what. Then, copy one tree block (e.g. “tree PAUP_1 = …..) into the other so that there is only one treeblock
  24. In TREES go to GET TREES FROM FILE
  25. Open HK TEST treefile and press GET TREES. Multiple trees should be imported into PAUP.
  26. Go to TREES, TREES SCORES, and click LIKELIHOOD
  27. Click TOPOLOGY TESTS
  28. Click Kishino-Hasegawa Test, Normal distribution, and one-tailed and hit OK