davidf at compusmart.ab.ca
Mon Mar 4 10:40:58 EST 1996
Kiyoshi Gotow wrote:
> Please try the following commands:
> The command "j" which tells the program to apply the JTT model is not
> always necessary, since it is the default setting.
> You can read the .EXA file by "SimpleText" or any other text editor.
> Good luck!
> Kiyoshi Gotow, Ph.D.
> Department of Human Life and Culture
> Junior College, Seitoku University
> gotow at vir.bekkoame.or.jp
Thank you very much for the advice. I was able to get the maximum
likelihood trees to work. Although, the exhuastive search was a bad
idea on my part, for it took almost 3 days to complete the search and I
ran into a memory problem when it finished. I eventually used the quick
option to complete my trees.
Now that I have results I am having a little trouble interpreting the
results. On most of my searches I was given more than one tree as the
most likely tree. There were two numbers that preceeded each one. What
do these numbers represent? I am hoping that they are probability
measures so i can accept or reject one of the trees with reasons other
than beacaused they fit the theory I want.
Also, I have noticed that when you run the distance calculation,
njdist, it gives you a distance matrix. This doesn't not appear to
fully scale the trees that protml produces, unlike the phylip dnaml
package. I am not all together familiar with the long hand statistical
clustering analysis that is used to come up with these results and I am
unable to scale my trees properly. Can you suggest anything for scaling
the trees that are produced in protml?
Thank You For Your Time
More information about the Bio-soft