Genetics: The computer revolution in progeny testing

Sunday, April 5, 2009

Thanks to the power of computers, we can now evaluate how progeny perform in the most challenging of commercial environments to drive higher rates of genetic and economic improvement

by SERGE POMMIER

In my last column, I briefly mentioned the role of advancing computer technology in establishing genetic markers and the positive impact these markers are having on the pace of genetic improvement.

Computers are also revolutionizing progeny testing. There's no better evidence of this than the emergence of crossbred breeding values.

For many years, genetics companies have had to rely primarily on information from nucleus herds when assessing the performance of pig lines. Evaluating genetic performance coming out of pristine nucleus herd environments did not always provide a full indication of how the genetics would perform in commercial farm operations.

With the power to collect and analyze millions of performance records and data, computers now allow breeders to factor environmental impacts or growing conditions into the selection process. Specific differences in environment, such as climate, nutrition, health, management and pre-slaughter handling, do not have the same effect on different genetic populations, or on the progeny of different individuals within these populations.

Genotype by environment interactions (often referred to as GxE) are particularly relevant to swine improvement programs as the progeny of individuals raised, tested and harvested under ideal genetic nucleus (GN) conditions have to perform under a range of varying commercial production and plant environments.

The key is to tie in the information obtained in multiple environments, using semen as the common link between the nucleus and the commercial environments. Hence, semen used at the nucleus level is also used at the commercial level in order to produce progeny which will yield additional information for the performance of that particular pedigreed animal.

Genetic correlations between purebred and crossbred performance can vary significantly. Research indicates that correlations for feed intake between purebred and crossbred performance can be as low as 0.50. In comparison, correlation estimates for carcass measurements or some reproductive traits can be as high as 0.75 to 0.80. This essentially supports the view that feed intake would benefit greatly from the added information coming from a GN crossbred program. Combining crossbred and purebred line selection methods has been shown to be superior to purebred line selection alone for certain traits – thus the emergence of GN crossbred programs.

Today, large networks of pedigreed breedings and progeny testing at the commercial level generate millions of records. And computers give us the power to process these records daily. So genetics that used to be evaluated based primarily on the performance of a few thousand sows in a genetic nucleus environment can now be tested using millions of pig records from around the world.

Computers also play a key roll in monitoring this massive amount of information and producing mating cards to control in-breeding.

This tremendous information capacity has allowed breeders to develop GN crossbred programs which increase the accuracy of crossbred selection by more than 30 per cent. And this translates into significant economic improvement. Economic modelling for GN crossbred programs suggests that the benefits could be as high as 19 per cent for the dam line and 26 per cent for the sire line over a five-year period.

Thanks to the power of computers, and their ability to process and analyze millions of performance records, we now have the capability to expand genetic line evaluation well beyond how a pig performs in high-health nucleus pyramids. We can now evaluate how the progeny perform in the most challenging of commercial environments to drive higher rates of genetic and economic improvement. BP

Serge Pommier, PhD, is director of technical operations for PIC in Canada.