Article,

On the Survival of a Gene in a Subdivided Population

.
Journal of Applied Probability, 3 (1): 142--155 (1966)

Abstract

A classical problem in population genetics consists in finding the probability that a line descending from a particular gene $A$ will become extinct. A simple assumption R. A. Fisher, Proc. Roy. Soc. Edinburgh 42 (1922), 321--341; J. B. S. Haldane, Proc. Cambridge Philos. Soc. 23 (1927), 838--844 is that of a large population where, initially, the number of $A$-genes is small. Another assumption R. A. Fisher, Proc. Roy. Soc. Edinburgh 50 (1930), 204--219; S. Wright, Genetics 16 (1931), 97--159 is that of a population with a constant number $N$ of genes where $A$ has a certain selective disadvantage over another gene $A$. The present author considers $K$ subpopulations which inhabit $K$ different ecological niches with possible migration between the niches. First, it is assumed that there is no limit to the size a population can attain and that the $A$-genes reproduce independently of each other. Let $\pi_i\ (i=1,2,\cdots,K)$ be the conditional probability that the line descending from an $A$-gene becomes extinct if the ancestral gene of the line is in niche $i$; the probability that $A$ survives is then expressible in terms of the $\pi_i$. Next, we assume that the population possesses a large but finite number $N_i$ of genes ($A$ or $A$) in niche $i$. If certain assumptions of independence are made, then the probability that $A$ survives is approximated by a function of the above $\pi_i$, provided these are not all equal to unity.

Tags

Users

  • @peter.ralph

Comments and Reviews