The development and application of
`genomic selection as a new
`breeding paradigm
`
`André Eggen
`Illumina Inc., San Diego, CA 92121
`
`Downloaded from https://academic.oup.com/af/article/2/1/10/4638585 by guest on 27 December 2021
`
`population will reach 9 billion people by 2030. Meeting the growing food
`need using fewer resources is therefore one of the greatest challenges that
`contemporary agriculture is facing. Recent estimations by the Food and
`Agriculture Organization of the United Nations (FAO, 2006) indicate that
`to meet the increasing demand, food production must double in the next
`50 years. Expressed another way, agriculture will have to produce more
`food than in the last 10,000 years combined. When the year 2000 is used
`as a base, projections indicate an increase in global meat consumption
`of 68% and in global milk consumption of 57% by 2030 (Steinfeld and
`Gerber, 2010). The greater demand for food based on animal proteins to-
`gether with the potential effects of climate change and water, nutrient, and
`energy scarcity will result in large productivity gaps. It is therefore critical
`to apply technical and scientifi c advancements systematically in feeding,
`nutrition, genetics, reproduction, animal health control, and general im-
`provement of animal husbandry to fi ll the coming productivity gaps. The
`largest gains will come from innovations that accelerate agriculture pro-
`ductivity while reducing costs and limiting environmental impacts (Cap-
`per, 2011).
`
`Traditional Breeding Methods Are
`Successful but Limited
`For centuries, animal breeders have very effectively manipulated the
`genomes of livestock species, making use of the fact that natural variations
`exist within a species, within a breed, and within a population. Traditional
`breeding has been done in the absence of molecular knowledge of the
`genes acting on a quantitative trait locus. Breeders have enhanced produc-
`tion traits in their herds by selecting superior individuals as progenitors
`for the next generations. These enhanced “breeding values” have been
`achieved by combining phenotypic recording of individual performance
`with genealogical information. In Holstein dairy cattle, milk production
`is still increasing by 110 kg per animal per year. In pig production, the
`kilograms of feed required to produce a kilogram of pork, known as feed
`conversion, is estimated to have decreased by 50% between the 1960s
`and 2005. Although these results illustrate powerful examples that can
`be achieved through traditional breeding methods, the effi ciency of these
`traditional methods decreases when traits are diffi cult to measure, have a
`low heritability, or cannot be quickly, inexpensively, and correctly mea-
`sured in a large number of animals. Such diffi cult-to-measure traits are
`often critically important; they include fertility, longevity, feed effi ciency,
`and disease resistance. Selection for these traits must be achieved through
`genomic means.
`
`© 2012 Eggen.
`doi:10.2527/af.2011-0027
`
`10
`
`Animal Frontiers
`
` •
`
` •
`
` •
`
`Implications
` A signifi cant gap exists between demand based on population
`growth and the current trajectory of yield; this is a point of lever-
`age for genomics.
` Although traditional breeding methods have been effective in
`selecting for animals with easy-to-measure production traits,
`these methods have essentially “hit an asymptote,” and more
`diffi cult-to-measure (and often more important) traits cannot ef-
`fectively be selected for using traditional methods.
` • The race to sequence the fi rst human genome, and, subsequently,
`the race to enable routine resequencing of tens, if not hundreds
`of thousands, of additional human genomes, has resulted in a
`100 million fold decrease in DNA sequencing prices since 1990.
`Animal genome sequencing has benefi tted from this.
`Sequencing and resequencing of economically important live-
`stock species has resulted in the discovery of millions upon mil-
`lions of single nucleotide polymorphisms. These single nucleo-
`tide polymorphisms are being deployed in massively parallel
`fashion on DNA microarrays, enabling genome-wide associa-
`tion studies to identify genotype-phenotype correlations for both
`simple and, more important, complex traits.
` • Driven by ever-increasing reductions in the cost of measuring
`genetic variation, we are entering a new era in which the in-
`formation from these genome-wide association studies will be
`utilized effectively in routine testing using genomic selection.
`Genomic selection holds promise for more widespread adoption
`than marker-assisted selection because it lacks the requirement
`of prior knowledge of alleles or marker positions of loci and the
`requirement that marker-assisted selection must be implemented
`within families.
`
`Key words: gene mapping, genomic breeding, genomic selection, next-
`generation sequencing, single nucleotide polymorphism
`
`Introduction
`In the past 50 years, the world has experienced an unprecedented in-
`crease in population growth. On the basis of current projections, the world
`
`Exhibit 1015
`Select Sires, et al. v. ABS Global
`
`

`

`Big Expectations of Genomics
`
`Table 1. Summary of first sequenced genomes for ani-
`mal species1
`
`Downloaded from https://academic.oup.com/af/article/2/1/10/4638585 by guest on 27 December 2021
`
`Species
`Chicken (Gallus gallus)
`Dog (Canis familiaris)
`Cattle (Bos taurus)
`Horse (Equus caballus)
`Pig (Sus scrofa)
`Sheep (Ovis aries)
`Cat (Felis catus)
`Rabbit (Oryctolagus cuniculus)
`Turkey (Meleagris gallopavo)
`Dromedary (Camelus dromedarius)
`Medaka (Oryzias latipes)
`Honeybee (Apis mellifera)
`1Modified from Fan et al. (2010).
`
`Genome size
`(assembly), Gb Year
`1.05
`2004
`2.4
`2003
`2.91
`2009
`2.47
`2009
`2.2
`2009
`2.78
`2008
`1.64
`2006
`2.67
`2009
`1.08
`2009
`2.2
`2011
`0.7
`2011
`0.236
`2011
`
`diseases; and further explain the control of complex characteristic fea-
`tures. It should accelerate the discovery process and help close the gap
`between genetics and the traits that are observed, known as the “genotype/
`phenotype gap.”
`With the production of whole-genome sequences for the major live-
`stock species at a fraction of the cost of the Human Genome Project, the
`comparison of sequences from several individuals of different breeds
`with a reference sequence resulted in an almost inexhaustible source
`of genetic markers, primarily polymorphisms in the form of single nu-
`cleotide polymorphisms (SNP). With the release of the fi rst draft of the
`chicken genome sequence (International Chicken Genome Sequencing
`Consortium, 2004), the chicken community was the fi rst animal com-
`munity not only to have access to millions of SNP, but also to be orga-
`nized in a freely accessible database, the Chicken Variation Database
`(ChickVD; http://chicken.genomics.org.cn/). Similar sequencing and
`resequencing efforts in several other livestock species resulted in the
`discovery of hundreds of thousands of SNP covering the entire genome
`(Ramos et al., 2009). These genetic markers were subsequently depos-
`ited in a publicly accessible database maintained by the National Center
`for Biotechnology Information, the dbSNP (http://www.ncbi.nlm.nih.
`gov/projects/SNP). This database currently contains SNP information
`for more than 86 organisms and a total of more than 50 million SNP. An-
`other major technological breakthrough has been the development and
`constant improvement of DNA array technology, which allows for the
`inexpensive measurement of SNP within a given sample. The success of
`these DNA arrays resides in the fact that they present a strong parallel-
`processing capacity, tremendous miniaturization, and a remarkable abil-
`ity to be automated. Although fi rst used for gene expression studies,
`these DNA arrays proved very useful for the development of whole-
`genome SNP panels for many species, including several agriculturally
`relevant species (Table 2). With DNA arrays, hundreds of thousands of
`SNP can be screened in parallel for the cost of a few hundred dollars, al-
`lowing scientists to perform genome-wide association studies that sim-
`ply would have been out of reach with traditional (i.e., microsatellite)
`markers. Over the past years, several studies have been published dem-
`
`Over the past 2 decades, the rapid development of genomics has opened
`new paths to address the scientifi c basis of livestock biology and breeding,
`and has resulted in new production methods to achieve sustained increases
`in animal feed yields and long-term improvements in the effi ciency of
`livestock production. A new era, the “genomics era,” promises to enable
`the objective prediction of consequences based on direct access to the full
`DNA sequence of many individuals, and therefore a renewed and more
`objective view of the genetic value of animals that is not limited to a few
`production traits.
`One of the triggering factors for the development of this genomic era
`was the international project to sequence the human genome (the Human
`Genome Project). The goal of this project was to produce the fi rst (de
`novo) full DNA sequence of a human being. Along with it came the devel-
`opment and implementation of new genomic tools, particularly improved
`DNA sequencing technologies and increased availability of high-through-
`put genotyping platforms. The price to sequence a single nucleotide of
`DNA has fallen 100 million fold since 1990. This is the equivalent of
`fi lling up your car with gas in 1998, waiting until 2011, and being able to
`drive to Jupiter and back twice. As Figure 1 illustrates, the rate of informa-
`tion coming from current-generation DNA sequencers is increasing expo-
`nentially, faster than the data from GenBank (http://www.ncbi.nlm.nih.
`gov/genbank/), faster than Moore’s law, and faster than other comparable
`high-growth scenarios. Technological breakthroughs that have driven this
`cost decrease have also facilitated the production of whole-genome se-
`quences for several animal species (Table 1).
`
`Genomics: Moving Animal Science
`to a New Dimension
`From a scientifi c point of view, accelerated genome-scale measure-
`ment will have a profound impact. It will fuel comprehension of the basic
`structure and function of livestock genomes; help unravel the history of
`life, characterizing the cause of relatively simple phenotypes and genetic
`
`Figure 1. The rate of information coming from current-generation DNA sequencers
`is increasing exponentially, faster than the data from GenBank (http://www.ncbi.
`nlm.nih.gov/genbank/), faster than Moore’s law, and faster than other comparable
`high-growth scenarios (http://www.ted.com/talks/lang/eng/richard_resnick_wel-
`come_to_the_genomic_revolution.html, courtesy of TEDxBoston 2011.
`
`January 2012, Vol. 2, No. 1
`
`11
`
`Exhibit 1015
`Select Sires, et al. v. ABS Global
`
`

`

`Downloaded from https://academic.oup.com/af/article/2/1/10/4638585 by guest on 27 December 2021
`
`Table 2. Currently available whole-genome single nucleotide polymorphism (SNP)
`chips developed for important agricultural species
`Species
`Identification1
`Classification
`Provider2
`Potato
`Potato
`Public
`Illumina
`Tomato
`Tomato
`Public
`Illumina
`Apple
`Apple
`Public
`Illumina
`Peach
`Peach
`Public
`Illumina
`Cherry
`Cherry
`Public
`Illumina
`Maize
`MaizeSNP50
`Commercial
`Illumina
`Rice
`Rice 44K
`Commercial
`Affymetrix
`Chicken
`Chicken
`Private: public sale
`Illumina
`
`SNP, no.
`8,303
`7,720
`8,788
`8,144
`5,696
`56,110
`44,100
`57,636
`
`Consortium
`SolCAP
`SolCAP
`RosBREED
`RosBREED
`RosBREED
`Commercial
`Commercial
`Cobb Vantress-Hendrix-
`USDA
`Morris Animal Foundation
`Neogen (GeneSeek)
`AgResearch
`Various
`Various
`Various
`Various
`Various
`Various
`Various
`
`Private: public sale
`Feline
`Cat
`Private: public sale
`Equine
`Horse
`Private: public sale
`Ovine
`Sheep
`Commercial
`BovineHD
`Cattle
`Commercial
`BovineSNP50v2
`Cattle
`Commercial
`BOS 1
`Cattle
`Commercial
`OvineSNP50
`Sheep
`Commercial
`BovineLD
`Cattle
`Commercial
`PorcineSNP60
`Pig
`Commercial
`CanineHD
`Dog
`1HD = high density; LD = low density.
`2Illumina Inc., San Diego, CA; Affymetrix, Santa Clara, CA.
`
`Illumina
`Illumina
`Illumina
`Illumina
`Illumina
`Affymetrix
`Illumina
`Illumina
`Illumina
`Illumina
`
`62,897
`65,157
`5,409
`777,962
`54,609
`648,000
`52,241
`6,909
`62,163
`173,662
`
`onstrating the effectiveness of using low-cost, whole-genome SNP ar-
`rays (e.g., BovineSNP50, OvineSNP50, EquineSNP50, PorcineSNP60)
`
` •
`
` •
`
` •
`
`For the fi ne-scale mapping of inherited defects when testing
`only a modest number of cases and controls (Becker et al.,
`2010; Brooks et al., 2010; Meyers et al., 2010).
`For the mapping of quantitative trait loci and DNA regions in-
`volved in the genetic mechanism of complex phenotypes, to dis-
`cover and estimate marker effects (Fortes et al., 2010; Minozzi
`et al., 2010; Pryce et al., 2010).
`For the characterization of population structure, to better docu-
`ment and understand the history of our livestock populations
`(Decker et al., 2009; Kijas et al., 2009; Gautier et al., 2010).
`
`The availability of whole-genome SNP panels is therefore bolster-
`ing the search for mutations underlying genetic variation in simple and
`complex traits and revolutionizing the speed at which gene regions and
`specifi c genes are being discovered. With the use of whole-genome SNP
`panels, the traditional approach, namely, positional cloning with a whole-
`genome scan to map the region of interest, followed by a time-consuming
`fi ne-mapping step, is fully replaced with an effi cient and cost-effective
`genotyping step using a whole-genome SNP array containing 50,000 to
`70,000 SNP.
`
`Genomics: A Paradigm Shift in Animal Breeding
`(Toward the Genome-Assisted Barnyard)
`The biggest (r)evolution is taking place in the application of genomics
`to the design and implementation of livestock breeding programs, prom-
`ising gains across the value chain. For breeders, breeding organizations,
`and members of the livestock industry, genomics is expected to increase
`the effi ciency and productivity of animal breeding, whereas for consum-
`ers and the processing sector, it should enhance security and the quality
`of animal products. New insights into the growth, nutrition, health, and
`protection of animals are expected, enabling a better understanding of the
`molecular mechanisms of traits of interest. Therefore, genomics proposes
`further opportunities to improve selection accuracy while decreasing the
`costs, reducing generation intervals, and exploiting new sources of poly-
`morphisms (Dekkers, 2004).
`
`From Marker-Assisted Selection to
`Genomic Selection
`Beginning in the 1990s, breeders used gene marker technology in
`the form of marker-assisted selection (MAS) to remove deleterious gene
`alleles (such as halothane in pigs or fi sh taint in brown-shelled chicken
`eggs) or to select favorable conditions based on some marker information
`(such as the Polled condition in cattle). The limitation of MAS is that it
`
`12
`
`Animal Frontiers
`
`Exhibit 1015
`Select Sires, et al. v. ABS Global
`
`

`

`Downloaded from https://academic.oup.com/af/article/2/1/10/4638585 by guest on 27 December 2021
`
`requires prior knowledge of gene alleles or markers that are associated
`with the traits of interest together with quantitative estimates of these as-
`sociations in the specifi c population. It must therefore be implemented
`within families. Furthermore, MAS explains only a limited part of the
`genetic differences between individuals. However, with the availability of
`cost-effective whole-genome SNP panels for the major livestock species,
`genomic selection tools are now leading the way to a paradigm shift in
`animal breeding. With suffi cient genetic markers, one can follow the seg-
`regation of the entire genome and not merely a set of specifi c regions of
`interest, moving from MAS to genomic selection. Parental relationships
`are no longer required to explain similar performances in animals because,
`with a dense set of SNP, similar performances can now be explained by
`the fact that animals are sharing identical chromosome fragments.
`
`Principle of Genomic Selection
`Genomic selection was fi rst described by Meuwissen et al. (2001) and
`is based on the fundamental principle that information from a large num-
`ber of markers could be used to estimate breeding values without having
`a precise knowledge of where specifi c genes are located on the genome.
`With tens of thousands of SNP, well chosen to be representative of the
`entire genome, it is expected that there will always be an SNP in close
`proximity to a particular gene or DNA fragment of interest; the exist-
`ing linkage disequilibrium between one (or several) SNP and a causal
`
`mutation will be substantial and can then be used to explain a signifi cant
`fraction of the variation of the observed trait. The fi rst step in the genomic
`selection process is therefore access to a large group of animals, either a
`reference or training population with accurate phenotypes for the trait(s).
`This population should also be genotyped using a whole-genome SNP ar-
`ray. The resulting data will then serve as a reference to develop a statistical
`model estimating the effect of each SNP with the trait(s) of interest. The
`result is a predictive equation to calculate a genomic estimated breeding
`value (GEBV). After a validation step, the genomic breeding value of new
`animals can be computed using the prediction equation, based on their
`genotypes from the SNP array and in the absence of any accurate pheno-
`types for these animals (Figure 2 and Table 3). The accuracy of the GEBV
`depends of the size of the population and the heritability of the trait to
`be considered. In the chicken, for example, González-Recio et al. (2009)
`referenced a 4-fold increase in GEBV accuracy over the parent average
`for feed conversion effi ciency.
`
`Implementation of Genomic Selection
`Genomic selection builds on existing breeding programs in which the
`collection of pedigree information together with phenotypic data is al-
`ready routine; it provides a new level of information that can be integrated
`into the decision-making process to identify and select the most promising
`animals. The principal advantages of genomic selection are that it can be
`
`Figure 2. A reference population of animals is scored for key production traits and genotyped using a commercial or custom single nucleotide polymorphism array. The
`genotypes are represented by the variable x, with values 0, 1, 2 (homozygous, heterozygous, or alternate homozygous). A prediction equation is generated, combining all
`the marker genotypes with their effects to compute a genomic estimated breeding value for each animal. This prediction equation can be applied to a group of animals that
`have not been phenotyped, breeding values can be estimated, and the best animals can be selected for breeding. Adapted from Goddard and Hayes (2009) by permission
`from Macmillan Publishers Ltd: Nature Reviews Genetics, © 2009.
`
`January 2012, Vol. 2, No. 1
`
`13
`
`Exhibit 1015
`Select Sires, et al. v. ABS Global
`
`

`

`Table 3. Example of a simplified calculation of the genomic breeding value with 4 single nucleotide polymor-
`phisms (SNP) and estimated effects (allele A vs. B) of +8, +4, +2, and −6 for SNP 1, 2, 3, and 4, respectively
`SNP 1
`SNP 2
`SNP 3
`SNP 4
`Genotype Value
`Genotype Value
`Genotype Value
`Genotype
`AA
`8
`BB
`−4
`AA
`2
`AA
`AA
`8
`AA
`4
`BB
`−2
`AB
`AB
`0
`AB
`0
`AB
`0
`BB
`AB
`0
`AB
`0
`AB
`0
`AA
`BB
`−8
`AA
`4
`AA
`2
`AA
`BB
`−8
`BB
`−4
`BB
`−2
`AB
`
`Animal
`1
`2
`3
`4
`5
`6
`
`Value
`−6
`0
`6
`−6
`−6
`0
`
`Genomic breeding value
`0
`10
`6
`−6
`−8
`-14
`
`Downloaded from https://academic.oup.com/af/article/2/1/10/4638585 by guest on 27 December 2021
`
`implemented very early in life, is not sex limited, and can be extended to
`any traits that are recorded in a reference population. It especially pro-
`vides for diffi cult-to-improve traits and better selection accuracy while
`reducing the generation interval, thereby increasing the intensity of selec-
`tion. It explains a much greater proportion of the genetic variance than
`MAS and, unlike MAS, is not limited to specifi c families.
`As early as 2006, Schaeffer (2006) showed that using genomic selec-
`tion, the genetic gain per year could be doubled in dairy cattle, with a
`potential to reduce costs for proving bulls by more than 90%. Rather than
`cattle going through a long and expensive progeny test with the recording
`of phenotypic information on large numbers of daughters, accurate GEBV
`could be calculated through a cost-effective genotyping step.
`In cattle, more than 15 countries are now using genomic breeding val-
`ues on the national level and have successfully passed the international
`GEBV test organized by Interbull (http://www.interbull.org/). The evident
`benefi ts observed in dairy cattle can also be translated to other species,
`and the broader use of genomic approaches, or even genomic selection,
`promises to be one of the next major advances in breeding programs for all
`animal species. As cost-effective genomic tools continue to be developed
`for livestock, crop, and aquaculture species, the corresponding breeding
`industries will be able to make selection decisions sooner, improve traits
`that are diffi cult to address with traditional breeding methods, and provide
`consumers with higher quality, safer foods while reducing the impact of
`breeding on the environment and ensuring its long-term sustainability.
`Applications of genomic approaches in the fi eld extend well beyond
`breeding because genomic tools can also provide pertinent and accurate
`information on animal identifi cation, validation of parental relationships
`(parentage), and traceability. Genomic approaches can be applied to the
`herd management process in optimizing mating and reducing inbreeding.
`
`Next Wave of Discoveries and New Approaches
`Higher density SNP arrays with several hundred thousand SNP are al-
`ready being developed in several livestock, crop, and companion animal spe-
`cies. In cattle, the success of genomic selection is being extended by combin-
`ing several breeds to increase the size of the reference population and perform
`across-breed evaluations. This approach should prove very benefi cial for
`breeds with a limited number of individuals or phenotypic records, or for spe-
`cies for which cross-breeding is an effective tool in the breeding process.
`As the cost of sequencing continues to decrease and access to the
`whole-genome sequence for specifi c individuals becomes affordable, one
`of the next steps will be to include whole-genome sequence data in routine
`genetic evaluations. According to a simulation presented by Meuwissen
`and Goddard (2010), a 40% gain in accuracy in predicting genetic values
`could be achieved by using sequencing data instead of data from 30,000
`
`SNP arrays alone. Furthermore, by using whole-genome sequencing data,
`the prediction of genetic value was able to remain accurate even when the
`training and evaluation data were 10 generations apart: observed accura-
`cies were similar to those in which the test and training data came from
`the same generation. According to the authors, “these results suggest that
`with a combination of genome sequence data, large sample sizes, and a
`statistical method that detects the polymorphisms that are informative . . .,
`high accuracy [in genetic/genomic prediction] is attainable” (Meuwissen
`and Goddard, 2010, p. 630).
`However, genomic selection today still treats the genome as a “black
`box.” It is not necessary to understand what is inside the black box to
`make effective selection decisions. From this point of view, it is therefore
`not much different from the broadly used and accepted best linear unbi-
`ased prediction method. Genomic selection could be further improved by
`integrating pertinent biological information and using effi cient methodol-
`ogies to get from the knowledge of a statistically associated marker locus
`to a functional gene variation. This would move from simply associating
`sequence variation with distinct phenotypes to a true understanding of the
`biology of the animal that makes those variations signifi cant, effectively
`shedding light on the black box.
`
`Opportunities for Developing Countries
`In developed countries, phenotypes and pedigrees have been recorded
`for certain species, such as dairy cattle, for more than 100 years. Progeny
`testing has been implemented for nearly 50 years. Developing countries
`are often limited by the absence of programs that record phenotypes on
`pedigreed animals and the lack of evaluation or national testing programs
`to assess the genetic value of germplasms. Genomic approaches should
`help in identifying critical populations for preservation together with
`some local well-adapted breeds that could be further utilized to breed
`valuable animals through a combination of selection and cross-breeding.
`Of course, as with genomics, you can manage only what you can measure,
`and collecting a minimum number of phenotypes in the fi eld will remain
`one of the critical and challenging steps to further deployment of genomic
`selection in developing countries.
`
`Conclusion and Perspectives for the Future
`The ability to investigate the genome, the transcriptome, the epig-
`enome, and the metagenome of any species by high-throughput sequenc-
`ing methods is opening a new world of possibilities. Further reduction in
`sequencing costs will continue to drive broader acceptance of new ap-
`proaches and their implementation for the benefi t of animal research, the
`breeding industry, and consumers. All economically impactful agricul-
`
`14
`
`Animal Frontiers
`
`Exhibit 1015
`Select Sires, et al. v. ABS Global
`
`

`

`Downloaded from https://academic.oup.com/af/article/2/1/10/4638585 by guest on 27 December 2021
`
`González-Recio, O., D. Gianola, G. J. Rosa, K. A. Weigel, and A. Kranis. 2009. Ge-
`nome-assisted prediction of a quantitative trait measured in parents and prog-
`eny: Application to food conversion rate in chickens. Genet. Sel. Evol. 41:3.
`International Chicken Genome Sequencing Consortium. 2004. Sequence and com-
`parative analysis of the chicken genome provide unique perspectives on verte-
`brate evolution. Nature 432:695–777.
`Kijas, J. W., D. Townley, B. P. Dalrymple, M. P. Heaton, J. F. Maddox, A. McGrath,
`P. Wilson, R. G. Ingersoll, R. McCulloch, S. McWilliam, D. Tang, J. R. McE-
`wan, N. Cockett, H. V. Oddy, F. W. Nicholas, and H. Raadsma. 2009. A genome
`wide survey of SNP variation reveals the genetic structure of sheep breeds.
`PLoS ONE 4:e4668.
`Meuwissen, T., and M. E. Goddard. 2010. Accurate prediction of genetic values for
`complex traits by whole-genome resequencing. Genetics 185:623–631.
`Meuwissen, T. H. L., B. J. Hayes, and M. E. Goddard. 2001. Prediction of total ge-
`netic value using genome-wide dense marker maps. Genetics 157:1819–1829.
`Meyers, S. N., T. G. McDaneld, S. L. Swist, B. M. Marron, D. J. Steffen, D.
`O’Toole, J. R. O’Connell, J. E. Beever, T. S. Sonstegard, and T. P. L. Smith.
`2010. A deletion mutation in bovine SLC4A2 is associated with osteopetrosis
`in Red Angus cattle. BMC Genomics 11:337.
`Minozzi, G., L. Buggiotti, A. Stella, F. Strozzi, M. Luini, and J. L. Williams. 2010.
`Genetic loci involved in antibody response to Mycobacterium avium ssp. para-
`tuberculosis in cattle. PLoS ONE 5:e11117.
`Pryce, J. E., S. Bolormaa, S. J. Chamberlain, P. J. Bowman, K. Savin, M. E. God-
`dard, and B. J. Hayes. 2010. A validated genome-wide association study in 2
`dairy cattle breeds for milk production and fertility traits using variable length
`haplotypes. J. Dairy Sci. 93:3331–3345.
`Ramos A. M., R. P. Crooijmans, N. A. Affara, A. J. Amaral, A. L. Archibald, J.
`E. Beever, C. Bendixen, C. Churcher, R. Clark, P. Dehais, M. S. Hansen, J.
`Hedegaard, Z. L. Hu, H. H. Kerstens, A. S. Law, H. J. Megens, D. Milan, D. J.
`Nonneman, G. A. Rohrer, M. F. Rothschild, T. P. L. Smith, R. D. Schnabel, C.
`P. VanTassell, J. F. Taylor, R. T. Wiedmann, L. B. Schook, and M. A. Groenen.
`2009. Design of a high density SNP genotyping assay in the pig using SNPs
`identifi ed and characterized by next generation sequencing technology. PLoS
`ONE 4:e6524.
`Schaeffer, L. R. 2006. Strategy for applying genome-wide selection in dairy cattle.
`J. Anim. Breed. Genet. 123:218–223.
`Steinfeld, H., and P. Gerber. 2010. Livestock production and the global environ-
`ment: Consume less or produce better? Proc. Natl. Acad. Sci. USA 107:18237–
`18238.
`
`About the Author
`André Eggen graduated from the Federal
`Institute of Technology (Animal Science)
`in Zürich (Switzerland) in 1988, where he
`also obtained his PhD in animal genetics in
`1992. Eggen then worked as a research sci-
`entist at INRA, where he became a research
`director in 2004. He was leader of the bovine
`genomics team, with research programs in
`the identifi cation of genes and chromosome
`regions for economically important traits
`in cattle, especially dairy quantitative trait
`loci and genetic disorders. He has more than
`100 publications in peer-reviewed journals,
`has participated in several European Union-
`funded projects and international consortia,
`and served as secretary of the International Society of Animal Genetics from
`2004 to 2010. Since May 2009, Eggen has been an agrogenomics specialist for
`Illumina, a global company that currently offers microarray-based products and
`services for an expanding range of genetic analysis sequencing. Illumina’s tech-
`nologies are used by a broad range of academic, government, pharmaceutical,
`biotechnology, and other leading institutions around the globe.
`Correspondence: aeggen@illumina.com 
`
`tural species, subspecies, and their pathogens will no doubt be sequenced
`in the near future. Thousands of related genomes will also be sequenced
`to sample genetic diversity within and between germplasm pools, offering
`critically important information for the implementation of genomic se-
`lection programs in developed countries. Genomic selection will surpass
`conventional methods as the dominant breeding paradigm, and specifi c
`haplotypes, detected via high-throughput genotyping or sequencing, will
`be directly associated with economic values. Breeding programs will be
`driven primarily by array data because of superior economics and much
`higher throughput. New expertise in the fi eld of animal pharmacogenom-
`ics will help increase vaccine and drug specifi city, whereas nutrigenomics
`will help tailor feeding regimens to genomic profi les.
`As genomic information continues to provide hugely valuable biologi-
`cal information, the key for further success of genomic selection and ge-
`nomic approaches will be to collect the most pertinent phenotypes, iden-
`tify the causal mutations and the exact mechanisms by which phenotypes
`are produced, and bring the different superior variants together in breed-
`ing lines in as few generations as possible. We are entering a truly exciting
`era fueled by genomics.
`Acknowledgments
`I acknowledge my colleagues Rob Cohen and Mike Thompson (Il-
`lumina Inc., San Diego, CA) for their critical reading of the manuscript
`and useful suggestions.
`
`Literature Cited
`Becker, D., J. Tetens, A. Brunner, D. Bürstel, M. Ganter, J. Kijas, and C. Drögemül-
`ler. 2010. Microphthalmia in Texel sheep is associated with a missense muta-
`tion in the paired-like homeodomain 3 (PITX3) gene. PLoS ONE 5:e8689.
`Brooks, S. A., N. Gabreski, D. Miller, A. Brisbin, H. E. Brown, C. Streeter, J. Me-
`zey, D. Cook, and D. F. Antczak. 2010. Whole-genome SNP association in the
`horse: Identifi cation of a deletion in myosin Va responsible for Lavender Foal
`Syndrome. PLoS Genet. 6:e1000909.
`Capper, J. L. 2011. Replacing rose-tinted spectacles with a high-powered micro-
`scope: The historical versus modern carbon footprint of animal agriculture.
`Anim. Front. 1(1):26–32.
`Decker, J. E., J. C. Pires, G. C. Conant, S. D. McKay, M. P. Heaton, K. Chen,
`A. Cooper, J. Vilkki, C. M. Seabury, A. R. Caetano, G. S. Johnson, R. A.
`Brenneman, O. Hanotte, L. S. Eggert, P. Wiener, J. J. Kim, K. S. Kim, T. S.
`Sonstegard, C. P. VanTassell, H. L. Neibergs, J. C. McEwan, R. Brauning, L.
`L. Coutinho, M. E. Babar, G. A. Wilson, M. C. McClure, M. M. Rolf, J. Kim,
`R. D. Schnabel, and J. F. Taylor. 2009. Resolving the evolution of extant and
`extinct ruminants with high-throughput phylogenomics. Proc. Natl. Acad. Sci.
`USA 106:18644–18649.
`Dekkers, J. C. 2004. Commercial application of marker- and gene-assisted selection
`in livestock: Strategies and lessons. J. Anim. Sci. 82(E-Suppl.):E313–E328.
`Fan, B., Z. Q. Du, D. M. Gorbach, and M. F. Rothschild. 2010. Development and
`application of high-density SNP arrays in genomic studies of domestic animals.
`Asian-australas. J. Anim. Sci. 23:833–847.
`FAO (Food and Agriculture Organization of the United Nations). 2006. World Ag-
`riculture: Towards 2030/2050—Interim Report. Food and Agriculture Organi-
`zation of