UNM Computer Science

Search Technical Reports by ID

The format of the tech reports ID number is TR-CS-YYYY-NN, where YYYY is the four digit year and NN is the number, including leading zeroes. For the first tech report of 2004, the search would be "TR-CS-2004-01".

This searches only by ID. If you'd like, you can also search by researcher or search by keyword

Found 1 result.

Listing from newest to oldest

TR-CS-2003-30

Controlled Breeding Problem
Tanya Y. Berger-Wolf, Cris Moore, and Jared Saia

We consider the biological problem of designing mating strategies for controlled animal breeding programs. Beginning with a small population of known individual genetics and or their pedigree, the aim of a breeding program is to come up with a mating strategy that creates and maintains certain characteristics (such as maximum genetic diversity) within the population. Typically, these strategies are derived based on extensive stochastic modeling, including computer modeling, based on statistical information about the population (birth rate, death rate, etc.). However, in most cases, especially in conservation biology, it is very hard to determine the population parameters. Additionally, in a small population, or a zoo population, one cannot apply statistical methods to derive or use these parameters. In this paper we propose the first discrete model of the controlled animal breeding problem and analyze two possible objectives: breeding for maximum diversity and breeding a target individual. The strategy design goal in both cases is to minimize the expected number of matings until the breeding objective is achieved. We evaluate several mating heuristics and provide upper and lower bounds for the expected number of matings. While the actual population parameters may vary and can change the actual number of matings for a particular strategy, order of magnitude of the number expected matings and the relative competitiveness of the mating heuristics remains the same. Hence, this first simple discrete model of the animal breeding problem provides a novel viable approach to designing and comparing breeding strategies in captive populations.

gzipped postscript