[Bio-software] Re: Phylip - WinXP

Joe Felsenstein via bio-soft%40net.bio.net (by joe At removethispart.gs.washington.edu)
Fri Dec 1 01:24:55 EST 2006


In article <mailman.80.1163788024.19683.bio-soft At net.bio.net>,
 <glykokalyx At hotmail.com> wrote:
>I got a problem with phylip running on winXP. Maybe someone has an
>idea. It would be appericated very much!
>
>Here comes the problem...
>After I have ran FITCH with my outfile from PROTDIS I get a tree-file
>of the size of 0kb. Without much surprise I can't open the file with
>the program Treeview. Of course, my original file contains sequences.

>Following a step-to-step explanation how I proceeded:
>1) Downloaded 3 protein-sequences from NCBI and saved them altogether
>in one single text file in FASTA format (>name [new line] sequence).
...
>After running PROTDIS my outfile-file looks like:
>    3
>putida      0.000000  0.713185  3.438244
>LB400       0.713185  0.000000  3.052331
>HD100       3.438244  3.052331  0.000000
>
>And after FITCH it looks like:
>
>   3 Populations
>
>Fitch-Margoliash method version 3.66
>
>                  __ __             2
>                  \  \   (Obs - Exp)
>Sum of squares =  /_ /_  ------------
>                                2
>                   i  j      Obs
>
>Negative branch lengths not allowed
...
>If anyone could give a helping hand, that would be great!

This is a "feature" of Fitch.  You get no result if you have only
three sequences.  You need at least 4.  My apologies for this
poor behavior.  However with three sequenceds there is only one
possible tree topology (a single interior node connected to all
three).  You can draw the tree by hand.  It is:

(putida:x,LB400:y,HD100:z);
 
putida      0.000000  0.713185  3.438244
>LB400       0.713185  0.000000  3.052331
>HD100       3.438244  3.052331  0.000000

>
where the three numbers x, y, and z are given by

x = (1/2)(0.713185+3.438244-3.052331) = 0.549549

y = (1/2)(0.713185+3.052331-3.438244) = 0.163636

z = (1/2)(3.438244+3.052331-0.713185) = 2.888695

If you run Neighbor you will get this tree, as Neighbor allows trees as
small as three species.  The fit is perfect in either case (as there are
three branch lengths being used to explain three distances.

----
Joe Felsenstein         joe At removethispart.gs.washington.edu
 Department of Genome Sciences and Department of Biology,
 University of Washington, Box 355065, Seattle, WA 98195-5065 USA



More information about the Bio-soft mailing list