Twenty five years ago livestock breeding companies were using the theory of multi-trait indexing as their latest tools for driving genetic improvement in their pig breeding lines. This means animals were individually measured for different traits.
Recorded physical measurements for these traits were fed into an index and the merit of the animals being tested would be expressed as a single number,” says Ed Sutcliffe, who achieved a distinction in his Masters in animal breeding in 1996 and has focused on developing and managing genetic programs for various pig breeding companies.
“This number would allow all pigs in a specific batch to be ranked according to merit, from which the best all-round pigs could be selected for further breeding,” says Sutcliffe. “The physical measurements (growth, backfat and feed intake) were expensive to gather and had to be weighted according to their economic production value, so the overall rate of genetic change was relatively slow.”
Best Linear Unbiased Prediction
However, the advent of the Best Linear Unbiased Prediction, BLUP, statistical theory, helped animal breeders up their game, commented Sutcliffe, who is currently technical director of UK pig breeding company ACMC. He also consults to independent breeders wanting to employ quantitative genetics within their own pig breeding programs.
“With the same physical measurements, the application of BLUP analysis meant that not only was selection more accurate (and therefore speed of genetic change increased), but we can compare pigs over time, and between environments,” says Sutcliffe. “We are no longer bound by the constraint of comparing animals in a common environment (at a given point in time).”
“We are making better use of the expensive physical measurements. Use of BLUP technology also opened up the ability to predict the genetic merit of offspring. While the accuracy of the prediction may not have been high, it was another tool to help pig breeders make selection decisions,” he said.
Pig DNA testing
In the early 1990s, molecular biology, through DNA testing, started to find a very important place in pig breeding.
“I should make it clear that I am not talking about direct genetic modification. The most famous example of DNA testing in pig breeding is the identification of the halothane gene as being directly related to Porcine Stress Syndrome, which is still widely used today. Much research was done but limited use was made when searching for these so-called DNA ‘markers’,” he explains.
“Once an association between a physical characteristic and a particular sequence of DNA at a known location was made, the DNA marker could be used to select for the physical characteristic without necessarily having to measure the pig for the trait itself,” says Sutcliffe. “These DNA markers are useful for selecting traits that are difficult to measure in the live pig, such as meat quality. The problem is that the association between the physical characteristic and DNA marker does not always hold true and can change over time. Also, including more than a couple of markers in selection decisions is not straight-forward.”
While Moore’s law (that computing power doubles every two years) may be commonly known, a similar phenomenon (exponential improvement) has been seen in the costs and speed associated with reading genotypes of animals from DNA samples.
“This has opened up yet more possibilities for animal and pig breeders,” says Sutcliffe. “Knowing the genotype of pigs at many locations along the chromosomes, and knowing the genotype of pigs in the pedigree of pigs, it is possible to track sections of chromosomes being inherited down the generations.”
“It is also possible, knowing the performance of these animals, to assign values to these segments of the animal’s genome as to how they influence a given quantitative trait such as growth, backfat and feed intake,” he said.
There was a lot of press coverage in the 1990s, for example, about breeding companies that were chasing the “holy grail” of being able to pluck a hair out of a new-born piglet and then confidently decide whether or not to select the animal for breeding. Sutcliffe points out that three or four years ago, it was also suggested that coupling genotype data to statistical analyses would allow a reduction in the collection of the expensive physical measurements.
Pig breeders physically measure a couple of generations and also genotype them to assess the “value” of the chromosome segments. Then they would simply genotype the next few generations and work out their merit by summing the “values” of each of the chromosome segments they carried.
“This would obviously appeal to breeding companies chasing the “holy grail.” They would not have to spend as much on performance trait measurements – but would have to spend money on genotyping more animals instead,” he says.
New hybrid approach
However, in keeping with the fast pace of technological development in the molecular biology field, this idea has largely been discarded and replaced by hybrid of BLUP and the molecular approach. Each pig is given a standard BLUP value and a new genomic value. Using the genomic value would roughly increase accuracy by 25% above BLUP evaluation alone.
“So, rather than reduce the importance (less animals being measured) of the phenotypic measurements, the current improvements in pig breeding are yet again extracting as much value as possible from the expensive phenotypes,” said Sutcliffe. “While the technology surrounding animal breeding develops a pace, there is one truism which persists.
“The cornerstone of any pig breed improvement program was, is, and always will be, good quality phenotypic measurements. While knowing the DNA sequence of a particular animal can tell breeders a lot, knowing how the “collection of genes” –the animal’s genome – actually equates to performance will always be more important,” he said.
As for the future of molecular biology in animal breeding – it is generally understood that the latest thinking will quickly be outdated. “It is thought likely that rather than reading the genotype at a few thousand places, we will be able to get the whole genome of animals for relatively little money in the foreseeable future,” said Sutcliffe.
“This will be like going from reading the first letter on every 14th page of a book with 840,000 pages to reading every letter of every word. Let’s hope Moore’s Law continues for many years to come if we are to make any sense out of all this data.”
In other words, pig farmers can look forward to a great deal of improvement on the genetic front.