junwen at astro.ocis.temple.edu
Thu Aug 21 19:11:55 CEST 2003
I am new in statistics, forgive me if this question is too silly. I want
to test if one alignment method is better over another one.
To do this, I use a bunch of sequence pairs as benchmark and aligned them
using these two method, respectively. By comparing the alignment with 3D
structure superimposed alignment, I can get a score (s1 for method one
and s0 for original method) for each alignment method on pairs I
What statistical test should I choose? I want to use a binomial test. but,
It seems this test only takes two set of data, either s1>s0 or s1<s0.
howver, in my case sometimes two methods have the same score, how to deal with the
situation when s0 = s1?
Or should I use paired t-test?
More information about the R-help