SwineFarming.org

Home › Articles › Swine Breeding Deep Dives › Genomic Selection in Swine: Current Adoption and ROI

Swine Breeding Deep Dives

Genomic Selection in Swine: Current Adoption and ROI

Genomic selection uses DNA markers, scored across thousands of positions in the genome, to improve the accuracy of breeding values for young pigs that have no performance records of their own. In swine it is now routine in the large genetic companies and is spreading to smaller nucleus and multiplier programs. This article explains how genomic selection works, where it pays, where the gains are modest, what adoption looks like, and how to judge whether the return on genotyping costs justifies the investment in your own breeding program.

How genomic selection works

Traditional selection ranks animals using estimated breeding values (EBVs) built from the animal's own records plus those of relatives. Genomic selection adds a genotype, usually from a SNP panel of tens of thousands of markers, so the model can see which chromosome segments an animal actually inherited from each parent. Full siblings share on average half their genes, but individual pigs differ a great deal in what they received. Markers capture that difference at birth.

The practical result is a more accurate EBV at a young age, before the pig has a litter of its own or any progeny. A tissue sample, often an ear notch taken at processing, is the only physical requirement.

Where the biggest gains appear

Gains are largest for traits that are expensive, late or impossible to measure on the selection candidate itself.

  • Sow traits in boars: litter size, born alive, and longevity can only be seen in daughters, so a boar candidate has no record of his own.
  • Low heritability traits: born alive and piglet survival have heritability often around 0.1 or less, so each extra bit of accuracy matters.
  • Carcass and feed traits: these require slaughter or feed intake stations, so genotype information from relatives raises accuracy on live selection candidates.
  • Disease-related traits: resilience measures that need challenge or field data are hard to record on the candidate.

For high-heritability traits already measured on every candidate, such as backfat, the extra gain is smaller.

Adoption in the industry

Adoption is deepest where breeding programs are large, centrally managed and well recorded. The major genetic suppliers genotype nucleus candidates as a matter of routine and use blended evaluations combining pedigree, phenotype and genotype. Adoption is thinner among small independent seedstock producers, where the barrier is less the lab fee than the lack of reliable pedigree and phenotype records to pair with the genotypes. Exact adoption rates vary by company and region and change quickly, so ask a supplier directly how their evaluation uses genomic information.

What genotyping really requires

Genotypes are only as valuable as the records behind them. A functioning program needs:

  1. Reliable parentage, since pedigree errors dilute accuracy.
  2. Consistent phenotype recording on a reference population.
  3. A genetic evaluation system able to combine all information; see genetic evaluation models for the BLUP versus single-step discussion.
  4. Tissue sampling and ID discipline at processing.
  5. Enough genotyped animals with phenotypes that the marker effects are estimated well.

A herd with sloppy ear tags and incomplete farrowing records will not be rescued by a SNP panel.

Estimating the return on investment

Return comes from faster genetic progress per year, usually by raising accuracy and sometimes shortening the generation interval, because young boars can be used earlier. Costs include genotyping per animal, data handling, sampling labor and evaluation fees. The benefit is realized mainly in nucleus herds and then passes down through multiplication to commercial sows, which is why genetic lag matters; see nucleus versus multiplication herds.

A rough approach: estimate the added annual gain in the index, convert it to dollars per market hog or per weaned pig using your supplier's economic weights, multiply by the animals that benefit, and compare it with total genotyping cost. Treat any figure that a vendor quotes as a claim to test with your own numbers.

Common mistakes

  • Genotyping everything without recording phenotypes alongside.
  • Ignoring pedigree errors and mislabeled samples.
  • Selecting hard on a single genomic trait and neglecting the balance covered in maternal line selection.
  • Expecting gains to show up in commercial herds within one generation.
  • Overlooking inbreeding, which can rise when genomic selection favors a few top families; see inbreeding depression in closed herds.

Who should and should not invest

Genomic selection suits a seedstock producer or breeding company with a stable recorded nucleus, a market for improved genetics, and access to a good evaluation. Commercial producers usually capture the value by buying genetics from suppliers who use it rather than genotyping themselves. A small producer buying a few boars per year gets more benefit from asking sellers about their evaluation and reading breeding stock selection than from running a private genotyping scheme. Targeted use on replacement gilts is sometimes worthwhile when a supplier offers it as a service.

Where it is heading

Expect lower genotyping costs, low-density panels with imputation, and more traits such as robustness and disease resilience. Pairing genomic information with the crossbreeding system in use will matter, because purebred selection must pay off in crossbred commercial performance.

FAQ

Does genomic selection replace performance testing?

No. Genomic predictions are calibrated against recorded phenotypes, so continued measurement of key traits is essential to keep predictions accurate.

How much faster is genetic progress with genomics?

It depends on the trait and the program, and published gains vary. The largest improvements are for low-heritability sow traits, but ask your supplier for evidence from their own program.

Can a small herd use it?

Yes, usually by buying genetics from a source that genotypes, or through a supplier service, rather than building its own evaluation.

Keep learning

Browse all 700 guides →