help for relative rate test

[yunwu zhang]

Dear Sir/Madam,
I am doing some work on gene evolution now and want to test rate constancy
using relative rate test. but because of poor mathematics, I still have
some questions about the method. First, how to give freedom degree, the
substitution between two sequences or the total number of nucleotides or
else? second, I am not certain about the relationship among variance,
corresponding variance and standard error, could any people please explain
that in detail? Formula will be favorable. 
Thanks a lot.
Yunwu Zhang