What is Genomic Selection?

Genomic Selection is the selection of animals based on genetic markers. When the markers effects are known and it is known which markers the animal carries, the breeding value of the animal based on these markers can be calculated and can be used for selection.

What is a genetic marker?

Genes are situated on the DNA and are responsible for the genetic differences between animals. For traits like milk production, hundreds of genes have an effect on the performance of an animal. It is unknown where exactly the genes that have an effect on specific traits are located and which variants of a gene (good or bad) the animal carries. However, we can follow the heredity of genes by the use of markers. Markers lying close to a gene have roughly the same heredity as the gene itself. CRV has identified 60,000 genetic markers that are being used to enhance the pedigree indexes of young bulls or the cow indexes of potential bull dams, which usually have low reliability. For animals with markers measured, it can be determined if the animal carries good genes or not. Animals that carry the ‘good’ genes on many locations get a higher ‘enhanced’ breeding value.

Can you tell me more about the reference populations?

For CRV, reference populations have been build for the Netherlands (only Holstein) and New Zealand (Holstein and Jersey) and CRV is building also a reference population for the USA. For young Holstein bulls from any country Genomic Breeding Values (GBV’s) on the Dutch and New Zealand scale can be calculated when the bull is genotyped.

For a reference population, genetic markers and BV’s are needed for hundreds to a few thousand bulls. The costs for genotyping a reference population are very high and therefore reference populations will become available for just a few countries.

Is there also a reference population for non-Holstein breeds?

With the ten thousands of markers that are currently used it is necessary to build up a reference population for every specific breed (results from Holstein can not be used for e.g. Jersey). This makes it harder to build a reference population for smaller breeds as Fleckvieh, MRY and Belgian Blue because these breeds have less AI tested bulls and it is more difficult to get a return on such large investments. When more markers or full genome sequences are available, smaller breeds can profit from the reference populations in the larger breeds because the functional mutations will be identified, or otherwise markers that are so close to the mutation that they will work across breeds.

Can you give me some examples of GS breeding values?

Example 1

Expectation values for the two full sibs (Jefferson x Wonderboy) born in 2007

 Milk-kg  Fat %  Protein %  Fat kg  Protein kg
 Parent average  +1557 -0.38  -0.14  +31  +40
 Genomic Selection
 Delta 2887  +1156  -0.10  +0.06  +36  +42
 Delta 2889  +696  +0.08  +0.11  +29  +30

Example 2

Expectation values and breeding values achieved for protein for Newhouse Sneeky and Newhouse Ricky.

 Protein kg (expectation value)  Protein kg (GS- expectation value)  Protein kg (GES)
 Newhouse Sneeky  5.3  23.7  29.0
 Newhouse Ricky  14.2  8.6  7.0

Example 3

Beeding values (parent average + GS-) and the GS effect for two full sibs (Paramount x Dustin) achieved for various traits

Paramount-B 1619 -0,31 -0,03 40 52 103 106 107 108 100 302
GS effect +156 -0,15 -0,02 -6 +4 -1,4 -1,2 -1,2 -1,3 -3,9 -44

 Milk kg  F-%  P-%  F-kg  P-kg  Frame  Udder  F&L  TS  Lgv
 Paramount x Dustin A  2212  -0.41  -0.13  55  64  107  111  110  113  391
 GS effect  +749  -0.25  -0.11  +10  +16  +2.8  +3.4  +1.2  +3.9  +46
 Paramount x Dustin B  1619  -0.31  -0.03  40  52  103  106  107  108  302
 GS Effect  +156  -0.15  -0.02  -9  +4  -1.4  -1.2  -1.2  -1.3  -44

The Paramount son A inherits 749 kg milk more (+2212) than expected based on his parent average. Paramount son B inherits feet & legs 1,2 points lower (107) than expected based on his parent average.

What is a genome?

The heritable information of an animal is stored in its DNA. The DNA is divided in chromosomes and all chromosomes together form the genome. A cow has got 29 chromosomes and 30.000 genes. From every gene are two copies (one from the mother and one from the father).

How do you get from marker information to genomic breeding values?

For Genomic Selection it is essential that genetic marker information and phenotypic information are available in large quantities. For phenotypic information it is most common to use breeding values (BV) of progeny tested bulls, but it is also possible to use phenotypic records of cows. With statistical analysis the association between markers and BVs can be estimated. When for a young bull it is known which markers the bull carries, his BV based on markers can be estimated.

Is the traditional genetic evaluation still needed when there is already so much information known about the bulls?

The traditional genetic evaluation is still needed because the estimated breeding values are used to calculate the marker effect on traits. It is also still needed because the breeding values of the bulls selected based on their markers still have to be estimated.

Are markers safe for the animal?

The animals are not genetically manipulated. The use of marker selection has no influence on the genes of an animal. Markers are a tool for selection at an early stage. This is done through blood testing. The integrity of an animal is not questioned with this technique. Natural genetic variation remains the source for genetic improvement.

What is going to change in CRV’s breeding program?

CRV used to select around 500 young black and white and red Holstein bulls, from which eventually 250 will be progeny tested. Now CRV sire analysts are sourcing around 1,000 young bulls. With this, the variation in pedigrees among the tested bulls increases and the extra information that Genomic Selection provides will highlight the qualities of new cow families more quickly and make it easier to utilize the strengths of these families. Out of these 1,000, with the help of Genomic Selection, the best 200 are identified to be progeny tested – a selection intensity of 1:5 compared to 1:2 previously – which leads to more genetic progress.

Delta Nucleus Programme Eurodonor Bulls:

  • Strong pre-selection in heifers via marker selection
  • Use maiden heifers stronger for OPU/ET
  • Higher value of embryo’s -Expansion of selection pool
  • Pre-screening bull dams via marker -selection
  • More calves per bull dam
  • Marker-assisted-selection of all calves
  • Young bulls at a higher level
  • Pass rate of bulls much higher
  • Broader offer: more segmentation
  • Marker selection of all calves

Do markers mean a revolution in cattle breeding?

Some expect radical changes in cattle breeding within 10 years. The current expensive system with test sires, lay-off sires and proven sires might then be history. If it will change quickly is hard to say although it is obvious that by using Genomic Selection soon after the birth of a promising young bull it can be determined what the genetic potential of that bull is. And using Genomic Selection on proven sires to enhance the reliability of their proofs will enable a more accurate choice of the right sires of sons to breed the next generation of bulls. On the female side, a more reliable choice can be made in the selection of bull dams.

What are noticeable changes because of Genomic Selection?

  • A higher percentage of young bulls will promote to a breeding bull. The young bulls will generally score better. There will be fewer disappointments.
  • Breeding organizations can use sire of sons at a younger age. After all, by the use of Genomic Selection we can find out much quicker which bulls are the most successful.
  • With Genomic Selection, CRV can anticipate quickly on changes and demands in the market (for example milk composition).
  • Genomic Selection makes product differentiation much easier. CRV can anticipate better on market segments. With GS, a wider product range can be achieved.

Since when is CRV studying Genomic Selection?

CRV has been investing in research into genes and genetic markers since 1994. Working with Belgium’s University of Liege and the New Zealand AI organization LIC, it originally looked into how far tracing specific genes could strengthen its breeding programme. After the acquisition of AmBreed, the cooperation with LIC was dissolved. CRV has been using Genomic Selection since 2006 and has invested heavily in selecting sires via genetic markers.

What is the difference between Marker Selection and Genomic Selection?

It is the same. In the communication of CRV both terms are used. The term Genomic Selection refers to the research method and Marker Selection indicates that the young bulls are selected based on their marker information.

What are the reliabilities of Genomic Selection?

A conventional tested animal’s parent average is around the 35% reliability. With Genomic Selection, we can increase this to more than 50%. The GS information roughly adds 15 percent extra reliability to the parent average of the bulls. This 15% extra reliability might increase when the reference population becomes larger or when more markers are available. Marketing GS bulls in a group will increase the reliability. For example, 5 bulls with a GS breeding value have together a reliability of 90%.

Thanks to Genomic Selection we know more about traits with a low heritability at an earlier stage, especially the durability and health traits. With using only daughter information, achieving a high reliability takes very long because we often need to wait until the daughters are culled. Genomic Selection is therefore a huge advantage.

What is the advantage for me as a dairy producer?

Markers give extra information, which is also available earlier. In the future, the breeding values of a bull will not only be based on the performance of his daughters but also on marker information. The breeding value estimation of bulls – particularly for the low heritability traits – will therefore be more reliable as more markers will be found. This is a clear advantage for dairy producers. CRV will also target its breeding programme towards different market segments. With Genomic Selection it is possible to predict – at a very young age – whether a bull will transmit the qualities desired by different markets. You can get tomorrow’s genetics today!


Our breeding goal is: ‘a healthy and long lasting dairy cow that will give an optimum contribution to farm profit’.


For the selection of bull dams and sires of sons, CRV in The Netherlands uses its own index. In this ‘CRV index’ a lot of weight is put upon health traits and longevity.

The relative weight of production, health and conformation in the CRV-index

 Production  Durability and health  Type
 40%  35%  25%


The NVI is the total merit index that is used in the Netherlands and Flanders for ranking bulls with the aim of putting those bulls at the top that are able to produce daughters that come closest to the national breeding goal.

In order to achieve the desired overall-breeding goal, economic weight is assigned to the production traits and weighting factors have been calculated for the functional traits such as longevity, health traits and conformation.

This overall breeding goal results in cows with a sustainable, efficient lifetime production. NVI is one of the best tools to help with that quest.

For relative traits, breeding values of sires higher than 100 are desirable, and for traits presented on the absolute scale breeding values higher than 0 are desired. This indicates improvement of the trait with respect to the previous generation.

Using NVI in your breeding programme will lead to:

  • Improvement of production and durability;
  • Lower culling rates, which means fewer replacements and thus a higher income;
  • More fertile and healthier cows, which will have a positive effect on income and consumer acceptance.


To achieve the goal set for the breeding programme, CRV in The Netherlands operates two separate breeding programmes: Delta and Euro Donor.


From over 1000 top ranked heifer calves, the best 100 heifers based on their genomic breeding values are selected for the Delta nucleus. These heifers produce many embryos that are sold to farmers with a buy-back contract for the resulting calf. At the end of pregnancy, the heifers move to one of five cooperating farms for lactation. The superior heifers based on production, type and functional traits will return to the Delta nucleus for additional embryo production.


Euro Donor selects the bull dams in The Netherlands, Flanders and the rest of Europe. After genotyping, the superior cows based on production, type and functional traits are contracted for embryos and bull calves. The embryo will be sold to farmers with a buy-back contract, similarly as the embryos harvested in the Delta program. From all bull calves genotyped, only the very best based on their genomic breeding values enter the progeny test program.


For specific pedigrees outside of Europe the US gene pool is a source of CRV genetics.