Path: utzoo!utgpu!watserv1!watmath!uunet!elroy.jpl.nasa.gov!usc!snorkelwacker.mit.edu!bloom-beacon!eru!hagbard!sunic!news.funet.fi!hydra!cc.helsinki.fi!aalto
From: aalto@cc.helsinki.fi (Erkki 'rkki' Aalto, University of Helsinki, Finland)
Newsgroups: bionet.population-bio
Subject: simulating random drift
Message-ID: <1991Mar5.195545.5429@cc.helsinki.fi>
Date: 5 Mar 91 19:55:45 GMT
Organization: University of Helsinki
Lines: 26


I am an ex-population geneticist working with computers for living.
I have been playing with an extremely naive population genetics model of
haploid organisms with discrete generations just for my own fun. This newsgroup
is not too crowded, so perhaps I am allowed to ask a simple question.

In the model the frequency of n:th allele in next generation
x_n(t+1)=F_nx_n(t)/N, where F is a rather complicated fitness function and N
is the normalization factor \sum_nF_nx_n. I would like to incorporate random
drift in the model, ie. replace x_n(t+1) with a random number with mean
X_n(t+1) and standard deviation some function of population size.

What would be the simplest way of doing this? A quick and dirty approximation
is enough, as I am after qualitative behaviour, and there is no realism in the
model anyway. I believe this should be a trivial question to anyone who has
done population genetics simulations. I have actually never done it, and I do
not have literature available.

Erkki Aalto

University of Helsinki

Finland

Internet: aalto@phcu.helsinki.fi
Bitnet: aalto@finuhcb