next up previous
Next: Likelihood versus Parsimony at Up: Phylogenetic trees Previous: Understanding Maximum Parsimony

Effect of Alignment on Tree-building

Here you will run parsimony on the 16S alignments that you made yesterday. The manual alignment carefully preserves the secondary structure that you saw in the exercise yesterday. Although there are ways to take this structure into account when making RNA alignments, clustal is not so good at this, which is good to know. Here you'll compare the manual alignment with the automated clustalw alignments and see their effect on trees. You'll also learn how to use the taxonomy browser at NCBI to explore species relationships. EMBOSS versions do not always work reliably, so download the source code and documentation for Phylip 3.6.

  1. Convert two 16S alignments to phylip format Popular software for making trees comes from the phylip package. EMBOSS comes with versions of some of the programs in the phylip package. We'll use one of these now. To do so we'll have to convert the alignments to ``phylip''file format. In annoying reality is that the phylip format is not fully standardized and different software reads and writes it in different ways.

    One useful conversion program on your system is sreformat. Execute

    % sreformat

    on the terminal to see how it works. Convert the 16S_manual.aln and one of your clustal/emma alignments of this file to phylip format with sreformat. Remember to save the converted data to file with a useful name like 16S_manual.plp.

  2. Run DNA parsimony on the two alignments. Find the relevant program with wossname. Then exeucte the program on your two alignments. Give the output files meaningful, distinct names.

  3. Download and install TreeView for Mac OS X. For this next part you'll need to look at the two trees. One good, but still incomplete, program for this is TreeView. Find and install it on your system.

  4. Use TreeView to compare the two trees. Which one makes better sense? You can necessarily tell this unless you know the meaning of the labels and know something about bacterial evolution. A large number of the bacteria in the tree are alpha-proteobacteria. Specifically these: (M.loti, S.meliloti, A.tumefaci, B.melitens, B.henselae,B.quintana). Many of the others are gamma-proteobacteria. One species that may have different locations in your two trees is Caulobacter crescentus, which is symbolized as ``C.crescent.'' You are going to find out which group this bacterium belongs now.

  5. Find the NCBI Taxonomy server on the net and search it for Caulobacter. What kind of bacteria is Caulobacter? (Hint: look at the links at the top of your search result). If your trees are different, which one is correct in its placement of Caulobacter?


next up previous
Next: Likelihood versus Parsimony at Up: Phylogenetic trees Previous: Understanding Maximum Parsimony
David Ardell 2005-01-28