1f. The development of genomic selection and application through optimal contribution theory in livestock breeding
Submitting InstitutionsUniversity of Edinburgh,
Unit of AssessmentAgriculture, Veterinary and Food Science
Summary Impact TypeTechnological
Research Subject Area(s)
Biological Sciences: Genetics
Agricultural and Veterinary Sciences: Animal Production
Summary of the impact
Impact: Economic: Genomic selection has revolutionised, and is now
standard practice, in the major dairy cattle, pig and chicken breeding
programmes, worldwide and provides multiple quantifiable benefits to
breeders, producers, consumers and animals.
Significance: Increased food production world-wide
Beneficiaries: Breeding companies, primary producers, consumers,
Attribution: Work led by Haley and Woolliams (Roslin Institute now
part of UoE).
Reach: Methodologies applied worldwide in livestock improvement,
and more recently applied in human genetics and plant breeding.
The key insight, first made by Prof. Haley (Group Leader, Roslin
Institute and UoE employed 1985-onwards) and senior postdoc Visscher
(Roslin institute employed 1995-2007) in 1998 [3.1] was the realisation
that dense single nucleotide polymorphism (SNP) arrays could be applied to
livestock improvement in several ways. Critically, as well as being used
to identify the locations of gene variants affecting phenotypes, SNP
arrays can also be used to make whole genome prediction of genetic merit
for complex traits (recently exemplified for height in humans [3.2], led
by Visscher). Based upon this original work, the application of whole
genome prediction to animal breeding was developed further by Meuwissen et
al. [3.3] and is now standard in dairy cattle breeding programmes.
Further, recent advances led by Prof. Woolliams (Group leader, Roslin
Institute and UoE, employed 1977-onwards) [3.4, 3.5] have guided
implementation in dairy cattle and more recently in the poultry industry.
Pig and beef industries now also implement this technology.
The basic concept of genomic selection is that a dense SNP chip can `tag'
every chromosomal `segment' of an individual, and with pre-calibrated data
the genetic merit of an animal for any measureable performance trait can
be predicted accurately. Potentially this revolutionises breeding
practices, allowing identification of genetically superior animals
essentially at birth. This makes breeding programmes more efficient, and
allows selection for traits that would otherwise pose huge logistical
difficulties, e.g. disease resistance, product quality or environmental
The greater the efficiency of a breeding programme, the greater the
danger of unintended side effects arising from erosion of genetic
variability, genetic bottlenecks and inbreeding. Research at Roslin
addressed and solved this problem, and provided the tools to implement the
solutions in practice. Specifically, optimal contribution theory,
developed and published by Prof. Woolliams in 1999 [3.6], provided the
theoretical basis to balance genetic selection and conservation of genetic
resources (including avoidance of inbreeding). This theory now has now
been extended by Woolliams and colleagues to include genomic selection
using dense SNP chips.
Through industrial collaboration, Roslin-led research (Woolliams)
produced software to evaluate and implement these solutions. All major
breeding programmes in the intensive breeding industries (pigs and
poultry) are now guided by this theory and its principles.
References to the research
3.2) Yang J.A., Benyamin, B., McEvoy, B.P., Gordon, S., Henders, A.K.,
Nyholt. D.R., Madden, P.A., Heath, A.C., Martin, N.G., Montgomery, G.W.,
Goddard, M.E. and Visscher, P.M. (2010). Common SNPs explain a large
proportion of the heritability for human height. Nature Genetics 42:
Details of the impact
The concept of genomic selection was first developed by Haley and
Roslin/SRUC have remained at the forefront of the development and
implementation of genomic selection. Optimal contribution theory
(described above), provided the theoretical basis to balance genetic
selection and conservation of genetic resources (including avoidance of
inbreeding) and has been adopted by breeding companies [5.1].
Applications of genomic selection currently have the greatest quantifiable
impact in the dairy cattle breeding, poultry and pig breeding industries.
Impacting dairy cattle
Dairy cattle breeding is a multi-million dollar international industry and
genomic selection has been implemented in large national breeding
programmes since 2008 (initially in the US, Canada, Australia, New
Zealand, Ireland and The Netherlands). It now has been extended to nearly
all major breeding programmes, including the UK (led by SRUC) in 2012.
When implemented in dairy cattle breeding, compared to traditional progeny
testing, genomic selection double rates of genetic progress whilst, at the
same time, reducing costs of the breeding programmes [5.2].
Currently, virtually all bull semen commercially available to dairy
farmers comes from bulls evaluated using genomic selection/prediction
techniques. However, cows born from genomically-selected bulls have
probably been milking for only 2 years in the REF impact window, in
countries that were earliest to adopt. Genetic trends for the dairy
populations in the US and Canada were analysed recently [5.3]. Attributing
only half of the recent additional genetic gains from the use of young
bulls to genomic selection — which is likely to be conservative — suggests
that genomic selection has contributed at least £200M to these dairy
industries in the last 2 years. This will rise rapidly as more of the
population descends from genomically-selected parents.
Poultry breeding achieves benefit through the extremely rapid
multiplication of stock. At Aviagen (an Edinburgh-based poultry breeding
company) a single male bird in the breeding nucleus today can be the
great-great-grandparent of ~50 million birds in 4 years' time, producing
ca. 76,000,000 kg of meat. Thus, small improvements in the precision with
which superior birds are identified and utilised in the breeding nucleus
scales up to huge benefits at the consumer level, after only four years
[5.2]. Conservatively, we may expect a genetic improvement in productivity
due to selective breeding of ca. 1% per year (as shown by 5.3), i.e. an
extra 760,000 kg from a single male bird after 4 years, with these
benefits cumulating across all male birds used in the breeding nucleus,
and also cumulating over years with each successful round of selection.
Increased precision of selection further increases these gains: Aviagen
implemented genomic selection in 2012 and an improvement of 20% in the
identification of genetically superior males would lead to a further
152,000 kg of meat per nucleus male breeding bird per year. Aviagen's
activities lead to 50% of the world's marketed chicken and >50% of the
Further, scientists controlling the Aviagen breeding programmes are
almost entirely Edinburgh trained, and The Roslin Institute currently
advises Aviagen on its breeding strategies and on optimal implementation
of genomic selection.
Optimal contribution theory has enabled efficient and sustainable breeding
programmes to be implemented in many species, and it is the key to
efficient long-term exploitation of both traditional and genomic
selection. Its current impact is best illustrated through interactions
with Genus, formerly PIC, which runs the world's largest pig breeding
Genus currently has a 36% share of the "technified" (improved genetics)
segment of global pork production, which comprises 40% of global pork
production. Therefore, Genus contributes the genetics for 14% of total
global pork production.
A LINK-funded project "Sustainable use of animal genetic resources"
(1999-2002) involving Roslin, SRUC, Genus and two other partners
identified optimal contribution theory as the single most potent
technology to bring an immediate clear improvement to the breeding system,
with a predicted 10-25% increase in genetic gain [5.4]. This extra gain
was assessed using software [5.5] developed through a joint Roslin
initiative, which is now widely used within the breeding industry. Genus
implemented these concepts through placement of a geneticist within the
Roslin Institute, from 2002 onwards.
Optimal contribution theory was first implemented into Genus breeding
programmes in May 2003, however the major benefits have arisen since its
use was scaled up after 2008. The company currently uses optimal
contribution theory to optimize selection and matings, on a weekly basis,
in 33 GN status lines from 10 farms on 4 continents. The data included in
these evaluations stretches back ~25 generations to the early 1980s. Over
13 million records are used in routine genetic evaluations, which are
performed weekly. Over 54 million piglets born alive are registered in the
Genus database, resulting from application of this technology,
contributing substantially to the world supply of pig meat, as indicated
by their global share described above.
Sources to corroborate the impact
5.1) David McLaren, Genus PLC. http://tinyurl.com/loqw5z6
5.2) Hayes BJ, Bowman, PJ Chamberlain, AJ Goddard, ME (2009). Invited
review: Genomic selection in dairy cattle: Progress and challenges.
Journal of Dairy Science 92, 433-443. http://dx.doi.org/10.3168/jds.2008-1646
5.3) "The impact of genomic selection on North American dairy cattle
breeding organizations", a presentation by J. Chesnais, et al. http://tinyurl.com/pg2vofj
5.4) Avendaño, S., Watson, K.A. and Kranis, A. (2010). Genomics in
poultry breeding — from utopias to deliverables. Proceedings of the 9th
World Congress on Genetics Applied to Livestock Production, 9:49. http://tinyurl.com/nfzayej
5.5) Van der Steen HAM, Prall GFW and Plastow GS (2005). Application of
genomics to the pork industry. Journal of Animal Science 83, E1-E8 http://tinyurl.com/oojjjn4
5.6) Rutten, MJM, Bijma P, Woolliams JAM and van Arendonk JAM. (2002).
SelAction: Software to Predict Selection Response and Rate of Inbreeding
in Livestock Breeding Programs. Journal of Heredity, 93, 456-458. http://dx.doi.org/10.1093/jhered/93.6.456