Genomic testing
Definition
Section titled “Definition”Genomic testing is genotyping an animal from a DNA sample (usually an ear tissue punch, hair or blood) and using the genotype to predict its genetic merit. In the US, genotypes go to the national evaluation (CDCB), which returns genomic predicted transmitting abilities (gPTAs) for production, health, fertility and type traits, and selection indexes such as Net Merit (NM$), Cheese Merit, Fluid Merit and Grazing Merit, or breed-association indexes.
Key terms:
- SNP panel: the set of single-nucleotide polymorphisms genotyped. Low- density panels (a few thousand SNPs) are imputed up to a standard density (around 50K SNPs or more) before evaluation.
- Reliability: how much of the true genetic merit the prediction captures. A young genotyped heifer has a much higher reliability than a parent-average estimate.
- Parent average (PA): the average of sire and dam PTAs, the estimate available without testing.
Uses on a dairy
Section titled “Uses on a dairy”- Ranking heifers within the herd, so that the top group is bred to sexed dairy semen, the middle to conventional, and the bottom to beef semen or sold.
- Parentage verification: genotypes confirm or correct the recorded sire, which improves records and evaluations.
- Recessives and haplotypes: identifying carriers of known lethal or fertility haplotypes to avoid carrier × carrier matings.
- Culling and selling surplus heifers early, before raising costs are spent.
- Embryo donors: choosing donors for OPU/IVF or flush programs.
How to pull it in DC305
Section titled “How to pull it in DC305”Genomic results are not calculated in DC305. They are imported from the genotyping lab or evaluation provider, usually as user-defined items (e.g. a genomic NM$, a genomic index, a “tested” flag). Item names and how they are loaded vary by herd; check the herd’s private dc305-vocab mapping.
Typical uses once loaded (item names illustrative):
LIST ID BDAT GNM FOR LACT=0 BY GNM heifers ranked by genomic NM$COUNT ID FOR LACT=0 GNM=0 heifers with no genomic value (untested)Typical range
Section titled “Typical range”Herd context: US Holsteins, freestall or dry-lot, genotyping heifer calves.
| Measure | Typical |
|---|---|
| Reliability of NM$, genotyped young heifer | about 65–75% |
| Reliability of NM$, parent average only | about 30–40% |
| Recorded-sire errors found by parentage check | about 5–15% of animals |
| Share of heifer calves tested (herds that test) | all, or a subset before breeding |
- Within-herd spread: NM$ across a herd’s heifer crop usually spans several hundred dollars, which is what makes ranking useful.
Common mistakes
Section titled “Common mistakes”- Treating gPTAs as exact. They are predictions with reliability below 100%; small differences between heifers are noise.
- Mixing evaluation runs. Values change at each CDCB evaluation and base change; rank within one run.
- Comparing indexes across breeds or bases. NM$ for Holsteins and Jerseys are not on the same scale.
- Testing without acting. The value comes from decisions (sexed vs beef vs sell); if every heifer is bred the same way, testing returns little.
- Ignoring parentage corrections and keeping wrong sire IDs in DC305.
- Ranking by one trait instead of the herd’s chosen index.
- Untested animals defaulting to 0. An empty genomic item exports as 0 and can rank untested heifers at the bottom (or middle) by mistake.