This server (see steps 1-3 below) is for detecting correlated mutations using the method described in:
Itamar Kass and
Amnon Horovitz (2002)
Mapping pathways of allosteric communication in GroEL by analysis of correlated mutations
Proteins: Struct. Funct. & Genet. 48, 611-617
This method which is based on the chi-square test does not take into account the evolutionary tree.
A more recent method we developed for detecting correlated mutations that reduces the
evolutionary noise by taking into account the evolutionary tree is described in:
Noivirt O, Eisenstein M, Horovitz A. (2005)
Detection and reduction of evolutionary noise in correlated mutation analysis.
Protein Eng. Des. Sel. 18, 247-253
The codes for this method can be downloaded from here.
A rule of thumb is that the results for any two positions i,j are significant if P < 0.05 unless:
(1) any of the Nn,Ex is less than one, or
(2) if more than 20% of the Nn,EX are less than five.
If P =< 0.001 the results are considered significant even if condition (2) is not met.
Last Updated on June 09, 2007
Server by Jaime Prilusky, 2002-2003
Bioinformatics and Biological Computing
Weizmann Institute of Science