jacknifing sites and estimating protein site-rate

[Joe Felsenstein]

In article <8rfqer$gvu$1 at mercury.hgmp.mrc.ac.uk>,
Andrew Roger  <aroger at is.dal.ca> wrote:
>> > 1) does anyone have/know of a script or small program that allows one to
>> > site-jacknife by randomly removing a user-defined number (or percentage)
>> > of positions in one's alignment.  Preferably something that can handle
>> > phylip sequential format?
[James McInerney:]
>> Does PHYLIP-Seqboot not do this?
>Yes Seqboot does this, but the last time I checked it does a mandatory 50%
>site removal. What I want is to be able to vary the number of sites that I
>can remove by percentage or simple numbers.  Trivial but frustrating for a
>non-programmer like me!!!

In SeqBoot version 3.5c change the lines starting at line 715 from:

      if (randum(seed) < 0.5)
        q = (newgroups - 1.0) / 2;
      else
        q = (newgroups + 1.0) / 2;
    } else
      q = newgroups / 2.0;


to, if you want to delete a fraction, say 0.4,

      if (randum(seed) < 0.4)
        q = (newgroups - 1.0) * 0.6;
      else
        q = (newgroups + 1.0) * 0.6;
    } else
      q = newgroups * 0.6;

(The 0.6 is, of course, there because it is 1 - 0.4).
I haven't tried this but it should work.

-- 
Joe Felsenstein         joe at genetics.washington.edu
 Dept. of Genetics, Univ. of Washington, Box 357360, Seattle, WA 98195-7360 USA


---