Genomic selection

Genetic selection programs have been the single most important contributor to the increase of efficiency of poultry and animal production around the globe. With pressure on food security building up through a growing and wealthier world population whilst resources are limited, once again the need for increased efficiency of animal production is paramount. New technologies derived from genomics are being put into place that hold high promises for another leap in progress by breeding programs.

In 2004 the genome assembly of the chicken was published (Hiller & many others), the first such assembly of a domestic animal. In a companion paper in the same issue of Nature (Wong & many others, 2004) the discovery of some 3 million SNPs (Single Nucleotide Polymorphisms, single point mutations to the genome), was reported. This opened the way to an entirely new approach of genetic selection that was proposed in a pivotal paper some three years earlier by Meuwissen et al. (2001). These authors suggested that if one was able to use so many genetic markers -tests which reveal DNA variation at a single location of the genome- that the entire genome could be covered at short enough distances, it would be possible to accurately predict the genetic potential of an individual on the basis of such so called genome wide markers alone. They called this application Genome Wide Marker Assisted Selection (GWMAS) or in brief Genomic Selection.

Since 2005 a true avalanche of research by geneticists in academia and industry has been directed at designing methods for application of the principle of GWMAS in breeding programs of domestic animals. One critical tool for application is the software that contains the algorithms to use the information of genotypes and trait measurements on a large reference population to derive the genomic breeding value of animals (selection candidates with unknown breeding value) which only have genotype information available. Several approaches have been developed and these are now available to commercial breeders.

At the same time high throughput genotyping technologies were developed which allow the genotyping of hundreds of animals for anywhere between some thousands and hundreds of thousands of SNPs in one day.

Application of genome wide selection has taken place from as early as 2009 when the first dairy cattle and poultry breeding companies implemented genomic breeding value estimations based on genotypes derived from 50 to 60 thousand SNPs.

The principal advantage of genomic selection is that the use of DNA markers allows the breeder to have an estimate of the breeding value of a selection candidate as soon as a DNA sample of the animal has been genotyped. This can be accomplished within weeks after the birth/hatch of the animal i.e. well before the animal is sexually mature and ready to be mated.

Since genetic progress is proportional to the accuracy of the estimate of the breeding value estimated on the selection candidates and inversely proportional to the time it takes for the selected animals to produce the next generation (the generation interval), the advantage of genomic selection is larger when generation intervals in the traditional breeding scheme are longer than the biological minimum and accuracies from traditional BLUP selection are lower.

Therefore most additional progress by genomic selection is to be gained when the breeding program focuses on reproduction traits. By definition these can only be evaluated long after sexual maturity in a traditional scheme and therefore such schemes traditionally work with extended generation intervals (as in layers and e.g. dairy cattle) or they accept low accuracies at the time of selection and mating (as in female lines of broiler and turkey breeding schemes and e.g. swine programs). For layer chickens as much as 50% increase of genetic progress beyond traditional breeding schemes can be obtained.

The impact of genomic selection on key efficiency traits in breeding programs for meat producing animals (broilers, turkeys and e.g. swine) should be lower in comparison since traditional breeding values for such traits are quite good at the time of mating. The selection candidates have passed the growing stage and therefore have a fairly accurate traditional breeding value based on their own performance and performance of their contemporary sibs.

Breeding for improved disease resistance and general robustness of animals has proven very difficult with traditional methods. An important reason for that failure is the simple fact that such traits are very lowly heritable i.e. the environment plays a dominant role and the genes do not matter so much.

However another problem with selection for such `difficult´ traits is that it is often impractical or expensive to measure these on the breeding populations in an accurate way. This is where genomic selection offers new hope: if such difficult traits are measured in a reference population that is also genotyped, this reference can then be used to estimate breeding values for such traits in all selection candidates by genotyping alone.

Genomic selection is a truly revolutionary breeding technology that creates enormous potential for increase of genetic progress for egg laying traits in poultry and opens up new options for genetic improvement of `hard to measure´ traits such as disease resistance and general robustness.

Geneticists have only just begun to discover all the options that this new technology offers. There is much more to come.

Hillier LW & many others. 2004. Sequence and comparative analysis of the chicken genome provide unique perspectives on vertebrate evolution. Nature. 432: 695-716

Meuwissen THE, Hayes, BJ & Goddard, ME. 2001. Prediction of total genetic value using genome wide        dense marker maps. Genetics. 157:1819-1829

Wong GKS & many others. 2004. A genetic variation map for chicken with 2.8 million single-nucleotide polymorphisms. Nature. 432: 717-722