Thank you Sergei for your comments.
I am analyzing a set of sequences of VSV from different parts of Mexico, typically is say that VSV evolves over positive selective pressure but I am finding that negative is maybe more important. Other authors have used this site to calculate dN/d/S (SNAP/README) based on Nei and Gojobori (1986) method and they have find high values of positive selection. If I analyze my data in that web server I get the same. Could you tell me some differences between the use of this method and your methods.

2.- Based on the results by your methods could I conclude that negative selection plays a more important roll that positive. (my sample size is 138 from 5 years)

Thank you
Sincerely,
Lauro.

Hi Lauro,

It seems that datamonkey points to evidence of episodic positive selection, although the power to locate individual sites is compromised by negative selection across the majority of the tree. This is merely a suggestion: currently implemented datamonkey tests do not directly test for which sites are under episodic selection.

Sergei