Calculating Evolutionary Distance, Part II

This is the last exercise. We'll use a web-based version of the Phylip program protist that uses substitution matrices to estimate evolutionary distances directly. This method is the first step that goes into estimating a ``distance matrix'' between sequences, from which we can estimate a complete evolutionary tree. In other words, for instance, we could use this method to ``rediscover'' the tree that we used to make the sequences with seq-gen, even if it was more complicated then the one we used.

We'll also use it to estimate the distances for the raw data.

Don't worry, you won't have to compile anything! We're being lazy now, just submitting our data over the web to France.

1. Go to the web-interface for Protdist. Point your browser to the The Pasteur Institute's front-end for Protdist.
   
   
2. Click on the Advanced protdist form near the bottom of the window.
   
   
3. Fill in your email.
   
   
4. Browse in your simulated protein data file. Use aa05.
   
   
5. Check the options and run. Notice that the default distance model is JTT. Click ``No'' on the ``Gamma distribution of rates'' option. Leave everything else alone and hit ``Run Protdist'' at the top.
   
   
6. Get the results. When it has finished executing. You'll see a page that starts with ``Results.'' Click on outfile to see a distance matrix. How does the distance compare to the % Change and the Poisson distance? Fill in the value in a new column in your table. Go back and take a click on ``protdist.out'' just to see what it says.
   
   
7. Repeat for aa2 and your real data. Finally, two more jobs and you are done. Repeat two more times for aa2 and your real multiple sequence alignment in ops_drome.aln. For the real data, find the entry in the distance matrix that corresponds to the first two Drosophila sequences you looked at before. Fill in the values in your table and compare.
   
   

Congrats! You're done!