Reviews

Animal domestication in the era

of ancient genomics
Laurent A. F. Frantz   1 ✉, Daniel G. Bradley   2, Greger Larson   3 and
Ludovic Orlando   4,5 ✉
Q1 Abstract | The domestication of animals led to a major shift in human subsistence patterns,


from a hunter–gatherer to a sedentary agricultural lifestyle, which ultimately resulted in the

development of complex societies. Over the past 15,000 years, the phenotype and genotype
of multiple animal species, such as dogs, pigs, sheep, goats, cattle and horses, have been
substantially altered during their adaptation to the human niche. Recent methodological
innovations, such as improved ancient DNA extraction methods and next-​generation
sequencing, have enabled the sequencing of whole ancient genomes. These genomes have
helped reconstruct the process by which animals entered into domestic relationships with
humans and were subjected to novel selection pressures. Here, we discuss and update key
concepts in animal domestication in light of recent contributions from ancient genomics.

Animal domestication was one of the most important their wild ancestors in the early stages of domestication
transitions in human history1,2, beginning with the are osteologically invisible.
long-​term association between hunter–gatherers and The second approach makes use of genetic data
wolves more than 15,000 years ago3. Following the emer- obtained from modern domestic animals to both
gence of mixed-​crop farming societies1,2, between 11,000 retrace their geographic and temporal origins13–17 and
and 4,000 years ago (roughly the  Neolithic through to discover the genetic basis underlying domestic traits18,19.
the Bronze Age) numerous other species became incorpo- For example, geographic patterns of genetic diversity
1
School of Biological and
rated within human societies, including, but not limited have typically been used to infer the location of initial
Chemical Sciences, Queen
Mary University of London,
to, sheep, goats, cattle, pigs, chickens and horses. Since domestication centres20–22. In addition, the substantial
London, UK. their domestication, animals have occupied a wide range phenotypic diversity, both between and within modern
2
Smurfit Institute of Genetics, of roles, from simply being tolerated, to being venerated domesticated animal populations, combined with pow-
Trinity College Dublin, within ritual practices, to providing humans with other erful techniques that make use of genetic variation at
Dublin, Ireland. benefits, including food, clothing, material for construc- the genome scale have revealed how specific loci affect
3
The Palaeogenomics & tion, transportation, herding and hunting. The diversifi- traits such as coat colour, size, fat content, circadian
Q2 Bio-​Archaeology Research cation of phenotypes, evident in multiple domesticated clocks and behaviour19. However, genomic information
Q3 Network, Research taxa, has also provided generations of biologists with a obtained from living animals provides only a contempo-
Laboratory for Archaeology
Q4 key model with which to study evolution4,5.


rary snapshot of a long-​term evolutionary process, and
and History of Art, The 


University of Oxford, The process by which humans voluntarily or invol- the validity of inferences about the past that are based
Oxford, UK. untarily transformed animals into the diverse resources solely on analyses of modern populations is contentious.
4
Laboratoire d’Anthropobiologie they now represent has traditionally been documented In the past decade, novel molecular techniques
Moléculaire et d’Imagerie de through two complementary approaches. The first have enabled access to genetic information from past
Synthèse, CNRS UMR 5288,
Université de Toulouse,
approach, based on the archaeological record, documents populations, offering the opportunity to combine
Université Paul Sabatier, morpho-​anatomical changes and cultural innovations the time-​depth of archaeology with the resolution of
Toulouse, France. through space and time6,7. Osteological changes8, age of genetic data (Box 1). Here, we synthesize how genetic
5
Lundbeck Foundation death and sex ratio profiles9, isotopic signatures10 and data retrieved from ancient animal remains (Fig. 1) have
GeoGenetics Center, traces of material culture (for example, harnesses11 and revolutionized our understanding of the process of ani-
University of Copenhagen, corrals12) represent some of the diverse markers for the mal domestication, from the early stages to the most
Copenhagen, Denmark.
shift in the relationship between humans and animals, recent transformations that have resulted from modern
✉e-​mail: laurent.frantz@
which we now refer to as indicators of domestication. breeding practices. One emerging, overarching theme
qmul.ac.uk; ludovic.orlando@
univ-​tlse3.fr However, the archaeological record is fragmentary, and debunks the simplistic view that reproductive control
https://doi.org/10.1038/ many traits (such as colouration, docility and fecundity) and isolation from wild populations are common fea-
s41576-020-0225-0 that probably diverged in domestic animals relative to tures of animal domestication, revealing instead highly

Nature Reviews | Genetics

Reviews

Box 1 | Reconstructing population history using ancient genomes Four main methodological advances have contrib-
uted to the characterization of genome-​wide DNA time
Over the past decade, many tools have been developed, modified and tested for series for whole populations32. First, ever-​improving
reconstructing population history using ancient genomes. The first step in analysing NGS instruments provided the necessary throughput to
ancient genome data generated from a next-​generation sequencing platform is to align enable shotgun sequencing of ancient genomes, even in
the short reads to a reference genome. The most used program is Burrows–Wheeler
cases in which the fraction of DNA from environmen-
Aligner142, which has been extensively tested with ancient DNA143. Assembling ancient
genomes (instead of aligning to a reference) is often difficult given the highly tal microbial contaminants vastly exceeds that of the
fragmented nature of ancient DNA molecules, making ancient genomes fairly difficult so-​called  endogenous DNA. Second, capture techniques
to use when no reference genome is available from a closely related species (but see have helped focus sequencing efforts on predefined sets
refs144,145). Following short-​read alignment, the genotype of an ancient sample can of targets so that population data can be recovered for
be inferred by comparing the sequence of short reads with the reference genome. only a fraction of the cost33, albeit with the drawbacks
Single-​nucleotide polymorphisms (SNPs), which are the result of single point mutations, that only predefined variants are assayed and that the
are the variant type most commonly used to reconstruct population history. Other capture step could produce bias34. Third, the whole
types of ‘structural’ variation, such as copy number variation and insertion or deletion, molecular toolkit underlying DNA extraction (such
can also be inferred, although working with this type of variation can be challenging as bleaching and double digestion35) and DNA library
due to issues inherent to ancient DNA (for example, small molecules).
construction (especially single-​stranded DNA meth-
The genotype of an ancient genome is then compared with those of other ancient and
modern genomes. One popular approach is to compare genotypes at SNPs that have been ods36,37) has been fine-​tuned to the challenging chemical
pre-​ascertained in the modern population146. This method minimizes the incorporation of nature of ancient DNA molecules. Last, these technical
erroneous SNPs arising from ancient DNA damage (such as deamination) in downstream developments, such as uracil–DNA glycosylase (UDG)
analyses, although it can also introduce biases34. Many population genomics tools exist to treatment (to remove C/G → T/A misincorporations
infer the population structure by comparing genotypes at SNPs in multiple ancient and from ancient DNA molecules)38,39, combined with the
modern samples, some of which have been tailor-​made for ancient DNA. Principal discovery that specific bone elements have generally
component analysis is one of the most commonly used tools to infer ancestry147 and can improved DNA preservation rates, especially the inner
accommodate highly degraded and low-​coverage sequence data by allowing an ancient ear bone40, have considerably enhanced experimental
genome to be ‘projected’ onto principal component axes. Other tools, such as the model- success. Altogether, these methodological advances
​based clustering method ADMIXTURE148, can also provide information about population
have provided access to a whole new range of samples
structure and admixture in a set of modern and ancient samples. Additional methods,
based on asymmetry in gene trees (D statistics149) or allele frequency correlations with extremely limited preservation that were previously
(F statistics146) as well as more explicit graph testing approaches, such as TreeMix150 incompatible with DNA analyses (for example, many
or AdmixtureGraph146, are robust tools for testing the existence of gene flow between samples from the Fertile Crescent41–43).
populations.
Early domestication
The geographic and temporal origin of domesticated
dynamic, non-​linear and taxon-​specific evolutionary

animals Q15


processes (Fig. 2). An understanding of the early phases of animal domes-

Neolithic tication has eluded biologists and archaeologists for
An archaeological period that Methods in ancient genomics decades. Aside from species that were unequivocally
began ~12,000 years ago in
The first successful extraction of ancient DNA was car- domesticated from geographically restricted wild ances-
the Near East (later in other
parts of the world), following ried out in the mid-1980s, with the cloning and sequenc- tors (for example, sheep and goats44) or that originated
the appearance of farming ing of short mitochondrial DNA (mtDNA) fragments in well-​defined domestication centres such as the Fertile
communities and the obtained from nineteenth-​century museum skin samples Crescent (for example, pigs45 and Bos taurus cattle46,47),
domestication of plants and of the quagga, a now-​extinct zebra species23. The PCR was the geographic and temporal origins of other species
animals. This period marks the
latest stage of the ‘Stone Age’
instrumental in overcoming the generally limited amount remain contentious. For example, progress in pinpoint-
and ends with the development of DNA in ancient remains, and to generate sufficient ing the origin of dogs has been hampered by a lack of
of metallurgy (Bronze Age). numbers of DNA copies from preselected loci. However, reference data from extinct Pleistocene wild canids, sub-
the amplification process was not without limitations, tle morphological changes between wild and domestic
Bronze Age
owing to the poor chemical nature of ancient DNA, populations during early phases of domestication and the
An archaeological period that
began over 5,000 years ago in which introduced recombining artefacts and inflated absence of unequivocal material culture accompanying
Southern Europe and part of error rates24. With the development of methods for DNA the early stages of domestication48,49. In the case of horses,
the Near East. This period is extraction from bones and teeth25, a considerably larger substantial morphological changes that differentiate wild
associated with the use of archaeological record became amenable to DNA sequenc- and domesticated populations of horses only appear in
bronze and, in some regions,
the advent of more urban
ing. However, until the advent of next-​generation sequencing the early Iron Age (~3,000 years ago), whereas the ear-
societies. (NGS), the amount of retrievable genetic information was liest evidence of harnessing, milking and corralling can
limited to mostly a handful of loci and specimens. With a be found in the Botai culture of the Eneolithic period
Next-​generation sequencing few exceptions26, pre-​NGS studies typically leveraged short in Central Asia, ~5,500 years ago12,50. A lack of strong
(NGS). Sequencing
stretches of mtDNA to retrace the phylogenetic affinities phylogeographic structure in horses and several other
technologies that allow
researchers to sequence entire between past and present species and populations27,28. As species has also hindered the power of modern genetics
genomes (DNA) or the ultrashort DNA fragments obtainable from ancient to retrace both the number and location of domestica-
transcriptomes (RNA) samples are naturally suited to the massively parallel tion centres. For example, the reciprocal monophyly of
substantially faster and sequencing capacity of NGS29, this technique rapidly led dogs and wolves implies that dogs are not genetically
cheaper than the older
technologies. Also known as
to the characterization of the first nearly-​complete ancient close to any specific wolf population, and that the wild
ultra-​high-​throughput genomes30,31 (see Fig. 1 for a summary of the resources ancestor of dogs is now extinct51,52. The latter is also true
sequencing. available for various domesticated animal species). in horses, where possible centres outside the traditional

www.nature.com/nrg
 Reviews

≥1× coverage Recent genomic evidence from ancient horses (from

4,700–4,000 years ago) has revealed the presence of a
now extinct lineage of horses in Iberia, ruling out this
population as the ancestor of modern domestic horses55.
200 It remains possible that these local horses underwent a

Number of genomes
domestication process analogous to the archaeologi-
cally and genetically attested process in Central Asia
that occurred ~5,500 years ago, and that this putative
domestic population in Iberia was subsequently replaced
by another domestic lineage. In fact, population replace-
100 ment is common in horses. The ancestry of the earliest
domestic horses associated with Central Asian cultures
of 5,500 years ago56 was replaced between 5,000 and
4,000 years ago by the ancestry associated with modern
horses55. Additional archaeological and genetic evidence
0 is required to test hypotheses related to independent
origins of regional populations.
<1× coverage The simultaneous analyses of ancient dog genomes
and the archaeological pattern of dog remains suggest
that modern dog populations may be derived from
independent wolf populations in Western and Eastern
200 Eurasia51, although this interpretation has been ques-
Number of genomes

tioned by some researchers57. The situation is much

clearer in North America, where ancient genomes have
unequivocally demonstrated that native American dogs
are the descendants of dogs that were introduced from
Siberia over 10,000 years ago58,59.
100
In species such as goats, for which the distribution
of their wild ancestors was geographically far more cir-
cumscribed, ancient genomes have revealed a complex
contribution of multiple, genetically divergent wild
lineages to early domestic populations42. This pattern
0 could be the result of either continuous incorporation
2013 2014 2015 2016 2017 2018 2019 of wild individuals into domestic populations or inde-
Year pendent processes of domestication in different regional
settings42, with the former scenario supported by the
Cattle Dog Goat Horse Pig fairly frequent incorporation of wild individuals into
domestic populations of cattle and pigs43,60,61. For exam-
Fig. 1 | Increase in the number of published ancient ple, evidence from ancient genomes demonstrates that
genomes for domesticated animal species. The local wild auroch populations contributed to the genetic
cumulative number of genomes (≥1-​fold coverage) or make-​up of domestic cattle lineages in North Africa,
genome-​wide data (<1-​fold coverage) that have been Europe and the British Isles41,62. In pigs, the Near Eastern
published each year since 2013 for five major domesticated genomic affinity of the first domestic pigs transported
species (cattle, dogs, goats, horses and pigs). The first ancient into Europe was nearly completely erased by gene flow
genome of a domesticated animal, the horse, was published with the European wild boar43,63. The genomic patterns
in 2013. Since then, the number of genomes published has
obtained from ancient dogs and wolves, however, sug-
Endogenous DNA increased rapidly , although many species, including
DNA extracted from the tissue chickens and sheep, have not yet been sequenced. To date, gest that dogs were almost entirely reproductively iso-
(such as bone or skin) of an genomics data for 451 specimens have been reported, lated from wolves in both Europe51,57 and the Americas58
organism that is no longer including whole-​genome sequences (≥1-​fold coverage) for for over 10,000 years, although limited gene flow likely
alive, whereas exogenous DNA occurred in specific lineages, such as arctic dogs64.
180 specimens and genome-​scale data (<1-​fold coverage)
originates from outside (such
as from soil bacteria). The
for 270 specimens. As the cost of sequencing goes down and
proportion of endogenous ancient DNA laboratory protocols become more efficient, Genetic integrity during domestication
DNA molecules can vary this number is expected to keep increasing in the future. Reproductive isolation. The prevailing narrative has held
considerably depending on the that animal domestication required the establishment
origin and the age of a sample.
domestication centre (the Central Asian steppes), such and maintenance of reproductive isolation between wild
as the Iberian peninsula, have been proposed on the and domestic populations (Fig. 2). However, recent inter- Q6


Capture techniques 

Also known as target basis of a higher genetic diversity in modern breeds53. pretations of the archaeological record have questioned
enrichment, these techniques By contrast, highly divergent genomes within cattle54 this assumption65, and the recently generated spatio-​
help focus sequencing efforts (taurus and zebu) and pigs (European and East Asian) temporal genomic patterns of pigs43,63, horses66, cows41
to a subset of the DNA
templates present in DNA
have been interpreted as evidence for the independ- and goats42 clearly demonstrate that severing gene flow
libraries, through hybridization ent domestication of geographically and genetically was not necessary to maintain domestic populations.
to target-​specific probes. divergent populations across Eurasia. Instead, the ancient genomic record suggests a new

Nature Reviews | Genetics

Reviews

a b
L L
A A
D D

L L
A A
D D

L L
A A
D D

L L
A A
D D

L L
A A
D D
Domestic Wild Domestic Wild

Fig. 2 | Influence of ancient genomics on models of animal domestication. a | The most commonly accepted
model of animal domestication prior to generation of ancient genomic data and recent theoretical advances based on
Q5 
zooarchaeology65,134. This model involves an initial founder effect that results in a strong demographic bottleneck during

early domestication (blue), followed by a dramatic demographic expansion (red) and multiple parallel founder effects
during breed formation (green). In this model, the mutational load (L) is expected to increase following domestication and
then again following breed formation, whereas genetic diversity (D) follows the opposite trend and the strength of
artificial selection (A) increases throughout history but is more pronounced during breed formation. b | A revised model of
animal domestication, in which gene flow from one or multiple wild populations has a prominent role and a limited initial
domestication bottleneck , followed by a dramatic demographic expansion and multiple parallel founder effects during
breed formation. In this model, the mutational load is expected to mostly increase following breed formation, whereas
genetic diversity decreases at first but then increases again as a result of introgression with wild populations (arrows).
The strength of artificial selection also increases throughout history but is more pronounced during breed formation,
as in part a. Bars are proportional to the overall levels of mutational load, artificial selection and genetic diversity in the
population. The effective population size is represented by the width of the phylogenetic lineages.

narrative for animal domestication in which persistent as the ‘domestication syndrome’. This syndrome includes
introgression with wild populations that did not feature various phenotypic traits, such as coat colouration and
in the initial domestication trajectory plays a promi- floppy ears, which appear in many domestic animals, as
nent part in the evolution of domestic animal genomic well as in foxes that were selected for tameness during the
ancestry (Fig. 2). Belyaev Fox Farm experiment69. Although the robustness
This new perspective raises puzzling questions as to of the domestication syndrome and the extent to which
how domesticated phenotypes can be maintained in the the Belyaev Fox Farm experiment could be used as evi-
absence of reproductive isolation. A current model based dence for its existence has been questioned70, the func-
on speciation theory proposes that only a limited num- tional enrichment of candidate genes under selection
ber of loci (termed an ‘island of domestication’) con- has generally lent support to the hypothesis that neural
tribute to phenotypic differentiation between wild and crest genes underlie some of the phenotypic differences
domesticated animal populations67. Interestingly, linkage between domestic and wild horses71 and dogs72.
Neural crest cell mapping studies in maize have identified a small number
A temporary cell that of loci that have a substantial effect on the morphological The founder effect. Another widely repeated assump-
differentiates into multiple cell
types involved in the formation
differences between maize and teosinte, suggesting that tion regarding domestication is that early reproductive
of the nervous component of such a model might also apply to plants68. In cases in isolation was associated with a founder effect, whereby
bones and cartilages. Research which a small number of genes explain major differences only a subset of the available genetic diversity in the
suggests that the behaviour between wild and domestic populations, selection needs wild population was incorporated into the domesticated
of neural crest cells may
only to act on these loci to maintain phenotypic integrity, population. Lower levels of genetic diversity observed
have been modified by
domestication, leading to the even in the presence of extensive gene flow. in modern domesticated animal populations have been
development of multiple traits Genomic work in pigs suggests that MC1R (encoding interpreted to support this claim73–75. However, it is
that are common across many melanocortin 1 receptor), the gene that underlies coat unclear whether this process is a by-​product of a domes-
domesticated animal species colour variation, may represent an island of domesti- tication bottleneck that took place during the early stages
(also known as ‘domestication
syndrome’), including
cation43. Genes involved in  neural crest cell formation, of domestication or a more recent restriction of genetic
depigmentation, smaller brain, migration and differentiation have also been proposed to diversity caused by nineteenth and twentieth-​century
floppy ear and shorter muzzle. underpin the common suite of biological features known breeding practices60. Interestingly, recent ancient plant

www.nature.com/nrg
 Reviews

Runs of homozygosity DNA data sets have demonstrated a decided lack of bot- well have sustained cattle husbandry in the Near East
(ROH). Regions of the genome tlenecks associated with domestication76. This finding is during the multi-​century droughts that took place during
that are depleted of consistent with recent animal studies that have tracked the so-​called 4.2k event41,85.
heterozygosity, which can arise genome-​wide patterns of genetic diversity through Earlier, Near Eastern goats also witnessed a homog-
when a diploid individual
inherits two identical stretches
time and revealed that, typically, the majority of genetic enization of their mitochondrial ancestry42. Neolithic
of DNA at a specific position of diversity was lost during recent centuries and not during populations possessed a strong phylogeographic struc-
the genome, due to the mating the early phases of domestication51,57. For example, in ture among Anatolian, Levantine and Iranian herds, ech-
of two closely related parents horses, individual genome heterozygosity dropped by oing separations among early farmers and implying the
(such as cousins). The length
~16% during the past 250 years of breed formation55. incorporation of a myriad of genetically differentiated
and the number of ROH across
the genome can provide
Similarly, in pigs, runs of homozygosity (ROH) are larger bezoar (the wild ancestor of domestic goats) populations
powerful information to infer and more frequent in wild animals than in domesticated during domestication. The correlation between geogra-
levels of inbreeding. animals, most likely as a result of overhunting in the wild phy and mitochondrial haplogroup eroded during the
and restocking with domesticated animals77. Modelling Chalcolithic period, which witnessed the initial spread
Artificial selection
The process by which humans
of ancient mtDNA data from cattle suggests that as of the now-​ubiquitous haplogroup A throughout the
breed animals to enhance few as 80 maternal founders precipitated the domesti- region and, ultimately, across the world42 (Fig. 3). The
specific characteristics (traits). cation of cows78 but this does not extend to the auto- mitochondrial ancestry of other species may also have
somal genome79, suggesting that such a domestication been affected during the Bronze Age. In dogs, for exam-
Yamnaya culture
bottleneck only affected females. ple, the introduction of canine haplogroup A in Europe
An early Bronze Age culture
from the northern shore of the
and the Near East was potentially driven by human
Black Sea (Pontic steppe). Later stages of domestication migrations from the steppes86 (Fig. 3). In pigs, European
In recent years, ancient genomic data have dramatically mtDNA haplogroups arrive in the Near East during the
Sintashta culture improved our understanding of the early domestication Late Bronze Age and early Iron Age87 (Fig. 3).
A Bronze Age culture of the
northern Eurasian steppe,
stages, yet many questions remain unresolved. Beyond Ancient animal genomics has also been used to infer
which is considered to be an early stages, these data sets also revolutionized our the time frame over which domesticated lineages were
offshoot of the Yamnaya understanding of later stages of domestication, includ- established. For example, material culture has provided
culture. ing processes such as trade and exchanges (Fig. 3), the evidence for canine harnessing in the Arctic as early as
dynamics of artificial selection and the process by which 9,000 years ago11. Ancient DNA from these dogs indi-
4.2k event
A severe aridification event
people managed animals (that is, animal husbandry) in cates that they belonged to the same genetic lineage as
beginning ~4,200 years ago, the past. modern Arctic dogs and that this lineage also gave rise to
which has been hypothesized the earliest native American dogs. This evidence suggests
to have caused the collapse of Migration, trade and exchange that sledge dogs may have been a key component for the
multiple civilizations across
Eurasia.
The Bronze Age was a period of substantial cultural initial peopling of the Americas58 (Fig. 3). However, the
and technological innovation that led to dramatic soci- dynamics of dog ancestry within the Americas is com-
etal changes in agricultural societies80. Ancient human plex, as multiple, genetically differentiated lineages of
genomics studies have revealed that these changes were dogs were introduced over the subsequent millennia
associated with long-​distance migration and concomi- by Alaskan Native groups and Inuit and European set-

Q9


tant shifts in the patterns of ancestral genomics through- tlers. Strikingly, this introduction of dogs with European
out Eurasia33,81–83. Ancient genomic data sets have also ancestry led to the replacement of the dog lineages that
demonstrated how this period has affected the ances- were introduced more than 10,000 years ago58.
try of multiple domestic species (Fig. 3). For example, Numerous additional historical processes that have
Q7 a major shift in horse genomic ancestry took place

affected domesticated animals over the past 1,500 years


between 4,100 and 5,000 years ago, a period that over- have also been uncovered using ancient genomics. For
laps with the timing of major human population move- example, the Muslim expansion was accompanied by
ments, including expansion of the Yamnaya culture and the spread of Sassanid Persian-​related horse ancestry to
the Sintashta culture56 across Eurasia. Interestingly, the both Europe and Central Asia, and changed the popu-
horse population associated with these cultures under- lation structure as well as the genetic make-​up of horses
went a substantial demographic expansion during this after the seventh to ninth centuries55. It is also likely to
period. Whether the spread of the Yamnaya culture or have contributed to the spread of zebu cattle genetics
the Sintashta culture in fact triggered or resulted from through the Sahel in Africa (Fig. 4). The introduction
this expansion requires further investigation. of pigs from China to the United Kingdom during the
The Bronze Age was also accompanied by striking Industrial Revolution as part of the breed improvement
shifts in the genomic ancestry of Near Eastern cattle process88 provides an interesting parallel89,90. Perhaps the
(B. taurus), which had remained stable during the previ- most striking example remains the worldwide spread of
ous six millennia41. This unprecedented genomic turno- numerous modern dog breeds that were established
ver reached a magnitude of up to 70% and was the result from European stocks during the Victorian era91,92.
of widespread introgression of zebu (Bos indicus) from
the Indus valley41 (Fig. 4). However, this introgression did Animal husbandry
not affect mtDNA ancestry, suggesting that the intro- The genomes of domesticated animals have been
gression was male-​mediated, and therefore promoted by shaped not only by global patterns of migration, trade
human herders. Interestingly, compared with B. taurus, and exchange, but also by innovations in husbandry
zebus are naturally adapted to dry environmental con- practices. Different management practices throughout
ditions84. The arrival of this novel genetic ancestry may history have had variable effects on the genomes of

Nature Reviews | Genetics

Reviews

Studbooks domesticated animals, as, generally speaking, genetic has long been suspected to have resulted in substan-
Registries that contain the list diversity decreases as the intensity of management tial drops in genetic diversity in various domesticated
of animals that belong to the increases. For example, the introduction of  studbooks animals. Recent work based on an almost continuous
same breed and for which the and the breeder’s equation after the Industrial Revolution genome time series of horses confirmed this hypothesis,
parents are known.

a
Breeder’s equation
A mathematical formula that
allows breeders to predict the
response to selection of a
specific heritable trait.
Siberia

Newgrange
? ?
Botai
Iberia
?
Fertile Indus Valley
Crescent

Horse Dog Pig Goat Cattle

Fig. 3 | Major dispersals of domesticated animals uncovered by ancient genomics. a | Pre-​Bronze Age dispersals.
Chronologically , the first dispersal shown on this map involves dogs and takes place from Siberia into the Americas, over
10,000 years ago. The second and third dispersals are nearly contemporaneous, involve dogs and pigs, and take place as a
result of the spread of farming from the Fertile Crescent into Europe, over 8,000 years ago. The fourth dispersal involves
goats originating in Western Anatolia and spreading into the Levant and Iran. The last dispersal is hypothetical and
involves the potential spread of Botai-​related horses ~5,000 years ago from Kazakhstan into the surrounding areas,
potentially as far as Eastern Europe and China. b | Post-​Bronze Age dispersals. Chronologically , the first dispersal shown
on this map involves zebu cattle (Bos indicus) originating in the Indus Valley and spreading to the Near East. The second,
hypothetical dispersal involves the potential spread of dogs during the expansion of Steppes cultures into Europe. The
third dispersal involves dogs and takes place as the result of the expansion of Thule culture (Inuit) into the American Arctic
and Greenland ~1,000 years ago. The fourth dispersal represents the dispersal of horses showing genetic affinity with
Sassanid Persian horses to Europe during the middle ages (approximately seventh to ninth centuries). The last dispersal
represents that of European dogs into the Americas following the discovery of the ‘New World’ in 1492.

www.nature.com/nrg
 Reviews

a as the overall diversity of horse genomes decreased sub-

80
stantially only in the past ~250 years55. A similar situ-
ation has been reported in dogs, although insufficient
ancient genomes have been generated to pinpoint the
60 exact timing of the decline in genetic diversity. Modern
dog breeds possess lower genetic diversity than both a
5,000-​year-​old dog genome from Newgrange, Ireland,
and modern village dogs, suggesting that modern
Admixture (%)

40 % zebu
breeding practices are at least partly responsible for this
decline51. However, pigs do not show any evidence of a
decline in genetic diversity over time, possibly owing to
20 long-​term gene flow with European wild boars43,63 and
the introgression of highly divergent Chinese popula-
tions into European pigs as part of the pig improvement
0 phase in the nineteenth century, which may have masked
the diversity loss resulting from recent demographic
bottlenecks in modern European breeds93. Charting the
diversity fluctuations through time in other domesti-
8,000 6,000 4,000 2,000 cated species is desirable and is now possible through
Years before present
the generation of time-​stamped ancient genomes.
Importantly, reduced levels of genetic diversity may
b have negative biological consequences, as genetic drift
is more prominent in small populations and reduces
the efficacy of purifying selection. Consequently, alleles
that would otherwise be purged from the population
have a greater chance of being maintained at higher
frequencies. Thus, even as overall genetic diversity
decreases in the population, both the number and the
proportion of  deleterious variants in a diploid genome
increase, inflating the mutational load. As a result, delete-
rious alleles have a greater likelihood of being expressed
phenotypically. Recent work in horses has provided
compelling evidence that the past 250 years of breeding
has increased the mutational load55, mostly as a result
of the development of purebred lines and the decline
of draft breeds following the mechanization of agricul-
ture94. Similar, or even higher, mutational loads may be
expected in other species, including dogs and cattle, in
which modern breeds are maintained from extremely
limited numbers of reproductive animals and which are
affected by recurrent genetic defects (for example, in
Holstein cattle95).
In natural conditions, high mutational loads in small
populations affect long-​term survival by reducing indi-
vidual fitness. Within human niches, the ramifications
of high mutational load for domesticated animals is not
as drastic as for wild animals, given human interference
Fig. 4 | Spatio-temporal pattern of admixture between cattle and zebu. The divergence and a greater degree of animal care. Additionally, domes-
between both the European and African cattle (Bos taurus) genomes and the South Asian ticated animals can have extremely high reproductive
zebu (Bos indicus) genome is substantial and predates domestication. Admixture between success (for example, elite bulls) despite having high
Q8 
these two lineages has been assessed both temporally and spatially. a | f4ratios estimating mutational loads. Therefore, the type of reproductive
zebu ancestry in ancient Near Eastern cattle genomes are plotted and show a fairly sudden management developed by breeders can relax selective
influx of zebu genetic ancestry from the Indus Valley region after thousands of years of constraints (relaxed selection) by partly decoupling muta-
stasis41, possibly driven by the 4.2k climate event that included multi-​century drought tional loads and fitness, thus leading to an increase in the
across the region. b | The current balance of ancestries assessed using multilocus frequencies of recessive lethal mutations96,97.
microsatellite variation. Sampled populations are indicated by points and the interpolated
Breeders also often select related individuals for
gradations vary from 100% B. indicus ancestry in their South Asian origin (darkest shading)
reproduction, which increases the probability of gener-
to purely B. taurus ancestry (lightest shading) in both Europe and within native breeds of
the forest regions of West Africa135,136. Zebu introgression was particularly important for ating offspring that carry homozygous genotypes. This
establishing the genetic make-​up of modern African cattle, and in many regions, such as inbreeding allows deleterious mutations to be phenotyp-
East Africa and the Sahel, this admixture was likely adaptive, as these zebu breeds perform ically expressed even when they are recessive, and can
better than taurine cattle in warmer, drier conditions. Part a is adapted with permission have dramatic biological consequences. Inbreeding can
from ref.41, AAAS. Part b is adapted with permission from ref.136, Wiley-​VCH. be measured in a single genome by identifying ROH,

Nature Reviews | Genetics

 Reviews

Purifying selection which can provide a powerful source of information for after chickens arrived in Europe, it was only heavily
Removal of deleterious variants how breeding practices affect the degree of relatedness selected for from ~1,000 years ago, coincident with a
in a population by natural in a population. Measuring ROH requires high-​quality major shift in the intensification of chicken production
selection. Also known as genotype calls that are only possible using high-​coverage that is also visible archaeologically across Europe102.
negative selection.
data. Levels of inbreeding have thus far only been esti- By contrast, other traits were selected for during
Deleterious variants mated in the 5,000-​year-​old dog from Newgrange, the the early stages of domestication. For example, a single
Alleles that have a detrimental genome of which was sequenced at 25-​fold coverage51. amino acid change in MC1R, the gene responsible for
effect on the phenotype of an The analysis of this genome revealed much smaller and black coat colour in pigs (Fig. 5), was already present in
individual.
fewer ROH than in modern dog breeds, suggesting that Anatolian pigs ~8,500 years ago, suggesting that this
Mutational load modern breeding practices, and especially modern breed trait was selected by early farmers43. A similar pattern
The mutational burden in a creation and maintenance, are more intense than during was reported in both goats42 and horses103, suggesting
population or an individual the Neolithic. that coat colouration was one of the earliest targets and
resulting from deleterious Additional information about past breeding prac- markers of domestic animals104.
variants.
tices can be gleaned by quantifying the number of This candidate gene approach is limited, given
Relaxed selection reproductive males and females in a population. This that the genetic basis of most domestic traits remains
The weakening or removal of a can be achieved by comparing levels of genetic diversity unknown. Overlaying functional annotations in the
selective pressure, such as between sex chromosomes, autosomes and mtDNA98. genome with signatures of selection, however, offers
Q10 when domesticated animals
In cattle, for example, gene flow from aurochs is evi- an alternative approach for identifying potential past


are less subject to selective 

pressure from predators.

dent in the autosomes but is absent in mtDNA41. This breeding targets. Multiple methods are available to
has been interpreted as a management strategy that identify selection footprints in the genome. The most
may have involved allowing insemination of domesti- common methods make use of the allele frequency
cated females by wild bulls41,99. In horses, a comparison differences among pairs of populations or the extent of
of the levels of diversity of the Y chromosome and the linkage disequilibrium within populations105. However,
autosomal chromosomes demonstrated that some cul- modern allele frequencies are the end product of a
tures allowed fewer males to breed and instead selected temporally dynamic process that can be retraced using
specific stallion bloodlines55. This male-​oriented breed- ancient DNA time series. This ancient DNA approach
ing strategy was not practised by the Romans, and only can provide allele frequency trajectories through time
became increasingly prominent in the past 1,000 years (Fig. 5) , which represent a powerful source of infor-
as a result of the growing influence of Oriental stallions mation to derive both the time and the intensity of
(Arabian, Persian and Turkmen)100. selection102,106,107.
The potential of these methods has been best demon-
Artificial selection strated in horses and goats, for which a range of different
Artificial selection is a hallmark of animal domes- selection targets have been identified in a wide range of
tication. Characterizing the archaeological context archaeological contexts. In horses, combined morpho-
underlying the emergence and spread of key traits is metrical and genomic evidence indicates that Scythian
crucial for our overall understanding of domestication. horse breeders from Central Asia targeted genes
Furthermore, identifying which traits were selected at involved in limb formation in their apparent effort to
the onset of domestication is essential for understanding select sturdier morphotypes ~2,500 years ago71. In addi-
the nature of the transition from wild populations into tion, the locomotory phenotype present in some modern
those that adapted to human niches. Archaeological data horse breeds, including ambling and speed, was selected
can be used to interpret the effects of the transition to only in the past 1,000 years55. In goats, selection scans of
domestication by analysing morpho-​anatomical traits, ancient genomes in Neolithic Iranian and Serbian pop-
such as size, shape and sex ratios. Ancient DNA data ulations revealed early (~8,000 years ago) signatures of
have the potential to identify a wide range of pheno- selection for genes implicated in coat colour, milking,
types and potential selection targets (as the relationship stature, reproduction and foddering42.
between people and animals intensified), as the genetic
basis of multiple phenotypic traits has been charac- Applications
terized in modern populations19. Ancient sequence Conservation
data can therefore be used to predict the phenotype An ancient horse genomics study showed that the
of long-​dead animals and, thus, provide insights into direct ancestor of Przewalski’s horse, until now consid-
the ways in which animals living in close proximity to ered the only true wild horse, was once a domesticated
humans were affected by the shift away from living out- horse56. These data sparked controversy concerning the
side the human niche. For example, some phenotypes conservation status of this flagship species. Is a horse
in chickens, such as reduced aggression and more fre- that became feral thousands of years ago even worth
quent egg laying, have been tracked through time using conserving? This question is rooted in the common
ancient DNA variation at the thyroid stimulating hor- view that there is an essentialist dichotomy between a
mone receptor (TSHR) locus101,102 (Fig. 5). These studies wild and a domesticated population. Yet, in the case of
demonstrated that a specific TSHR allele (a Gly558Arg Przewalski’s horse, should a species discovered in the
missense mutation), which is found at high frequencies wild in the nineteenth century and possessing multiple
in modern domestic populations, was not ubiquitous clear adaptations to its environment now be disregarded
in early chickens101. Additional analyses showed that as a conservation priority for the sole fact that it inter-
although the frequency of this TSHR allele was ~40% acted with humans some 5,000 years ago? This question

www.nature.com/nrg
 Reviews

a Chicken TSHR b Pig MC1R c Horse DMRT3

1.00 1.00 1.00

0.75 0.75 0.75

Allele frequency

0.50 0.50 0.50

0.25 0.25 0.25

0.00 0.00 0.00

30

50

00

0
0
00
00

rn

ild

00

50

00

rn

0
0
0
0
0
0
0
0
50 0
16 –850
13 –600
11 –350
Mo 100
rn
0

0
–80

–85

38 –310
36 –285
33 260
31 –235
28 210
26 –185
23 160
21 –135
18 10
de

de

de
–20

–20

–19

–17

–15

0–5
0–1

–75

–62

–30
0–w

–
Mo

Mo
00

00

00
50
00
–

–
80

00

70

00

80

90
50

00

00

00

00
50
00
50
00
50
00
50
00
00
11

25
22

21

19

19

19

83

73

45

41
10

Years before present Years before present Years before present

Fig. 5 | Frequency of alleles underlying modern phenotypes in key domestic animal species. a | The frequency of a
derived allele (a Gly558Arg missense mutation) of TSHR (encoding thyroid-​stimulating hormone receptor) in chickens.
This allele was first found to be under selection in modern chicken populations137 and is associated with phenotypes such
as longer incubation time (that is, development), fewer fearful and aggressive behaviours and decreased levels of thyroid
hormones138. Ancient DNA analysis101,102 indicates that the increase in the frequency of this TSHR allele was due to
selection, likely taking place during the Middle Ages in Europe. b | The frequency of a derived allele (an Asp124Asn
missense mutation) of MC1R (encoding melanocortin 1 receptor) in domesticated and wild pigs. This allele results in black
coat colour and loss of camouflage coat colour in pigs139. Ancient DNA work suggests that this allele was selected very
early during the domestication of pigs, as its frequency was already ~50% around 8,000 years ago, whereas it is almost
absent in the wild (~2%)43. c | The frequency of a derived allele (a substitution that introduces a premature stop codon;
Ser301STOP) in DMRT3 (encoding doublesex and mab3-​related transcription factor 3) in horses. This allele is associated
with the ability to perform gaits in horses (for example, ambling or pacing)140. Ancient DNA analysis suggests that this
allele likely first appeared in Europe during the Middle Ages141. As opposed to the example in parts a and b, this allele does
not become fixed (or nearly fixed in the case of pigs) in the population. This is because only few horse breeds have been
selected for the ability to perform gaits and the mutation is therefore still segregating in the worldwide horse population.

has deep implications for conservation biology as a Animal health

whole. For example, there are many other feral species Ancient DNA techniques provide access to genetic mate-
around the world, some of which have negative impacts rial not only from the host organism but also from their
on the environment (for example, pigs in Australia108) pathogens. Over the past 10 years, ancient pathogen
whereas others have neutral or even positive impacts genomics has illuminated the causes and consequences
on their ecosystems (for example, sheep in St Kilda109 of historical disease epidemics in humans, such as
and pigs in the Komodo Islands110). The question can the Justinianic plague and the Black Death119. Domesticated
be extended to multiple other scenarios, such as spe- animals also suffered many epidemics in the past, such
cies that have been translocated as wild into non-​native as the eighteenth-​century European rinderpest epi-
environments (for example, pigs in Cyprus111 and many demic that killed up to 80% of affected cattle120 or, more
other vertebrate species, such as deer or jungle fowl, in recently, Marek’s disease (twentieth to twenty-​first cen-
Island Southeast Asia112) or domestic populations that tury), which has been reported in over half the countries
Justinianic plague have strong cultural value (for example, rare endan- across the world121 and reached 30–60% mortality rate
A historical pandemic of gered breeds113,114) and that provide obvious, consid- in the 1960s122. In addition, domestication facilitated the
Yersinia pestis (541–542 AD) erable services to humans. It is also important to note emergence of zoonotic diseases with potentially sub-
that affected Mediterranean
port cities, including
that an increasing number of archaeological discoveries stantial effects on long-​term human and animal health.
Constantinople, and which are leading to a more nuanced view of environments Ancient pathogen genomics is therefore immensely
resulted in the death of 25–50 that are considered ‘pristine’, such as the Amazon promising in the context of domesticated animals, not
million people. basin115–117 and Kruger National Park118, which in fact only to reveal the history of epidemics in domesticated
have often been extensively modified by humans in animals and humans but also to improve our knowledge
Black Death
A historical pandemic of the past. By providing the temporal resolution nec- to fight against infectious diseases in the future. The
Yersinia pestis (1346–1353 essary to reconstruct the historical interplay between example of leopard spotted complex, which is associ-
AD) that resulted in the death human activities and the biosphere, the archaeological ated with colour night blindness in horses123, perfectly
of 75–200 million people sciences, including ancient DNA, have the potential to illustrates that ancient DNA data may equally be use-
across Eurasia and which is
thought to have had a
inform conservation priorities beyond the unneces- ful to understand the history of genetic disorders and
profound effect on European sary, reductive dichotomy between wild and domestic the processes by which they arose (for example, strong
history. populations. artificial selection).

Nature Reviews | Genetics

Reviews

Perspectives to map regulatory and functional elements onto the

Over the past decade, ancient genomics has begun to genome of economically important livestock128. In addi-
dramatically shift our understanding of the process tion, recent methodological developments in ancient
of domestication. Numerous studies have provided DNA research are providing the potential to type varia-
novel insights into the location, timing and subsequent tion beyond single-​nucleotide polymorphisms, includ-
human-​mediated transport of domestic animals, as well ing copy number variation, epigenetic markers and
as the ways in which domestic animals were selected faecal and oral microbiomes129. Insights gleaned from
and managed in the past. By documenting how live- these approaches will be crucial for our understanding
stock populations endured both past epidemics and of the recent evolution of complex traits.
environmental change, ancient genomics can provide The full potential of ancient genomics to generate
invaluable information that can be used to address insights into the pattern and process of domestication
current and future societal challenges. These insights has yet to be realized. For example, coat colour, sex
could prove instrumental in a world in which the farm- determination and genomic relatedness inferred using
ing industry increasingly depends on antibiotics and ancient DNA data can be leveraged to provide cru-
vaccination, genetic resources for domesticated animal cial information about the ways in which humans and
populations are becoming more depleted (for example, animals interacted (including ritual practices) in past
>30% of domesticated varieties are now endangered124) societies71. Ancient DNA data can also be used to track
and global warming is altering selection pressures. hybrid species, including camelid hybrids (for example,
Although a great deal of progress in understanding tulu) and equine hybrids (such as mules and hinnies130),
animal domestication has been made, several key ques- which often possessed desirable characteristics that
tions remain, especially in species for which ancient include enhanced immunity and stamina compared
genomic data have yet to be generated on a large scale, with the parental species. Insights such as these can be
including cats, sheep, camels, chickens, bees and various particularly useful when investigating the military and
microorganisms, including yeast. Even in well-​studied trade logistics of past societies, and can be supplemented
species, the available data have often raised more ques- by genetic information extracted from parchment131, tex-
tions than they have answered. This is particularly true tiles132 and artefacts133, to advance our knowledge of past
for dogs, for which the timing, location and context in economies.
which the domestication process was initiated remain Over the past few decades, sequencing of ancient
uncertain. In addition, whereas new insights have DNA has become firmly entrenched within the mod-
recently been made regarding how humans managed ern toolkit of bio-​archaeological research. Combined
animals in specific cultures, many cultures remain with data from isotope studies, morphometrics, pro-
to be investigated. Documenting where and when teomics and radiocarbon dating, and interpreted within
traits important for domestication first emerged, and the zooarchaeological context from which the remains
where and when they were selected, will enable the were recovered, ancient genomes are generating novel
reconstruction of a step-​by-​step biological history of and often surprising insights into both the pattern and
domesticated animals. process of domestication. These insights are facilitating
The genetic architecture of most phenotypic traits is a new perspective on the relationships that humans have
complex and species-​specific, as was demonstrated in had with animals over at least the past 15,000 years, and
modern rabbits, in which production traits are highly how these relationships have become ubiquitous. The
polygenic75. However, in some cases, a quantitative trait next decade will undoubtedly witness a surge in the vol-
can be predicted using only a handful of loci; for exam- ume and quality of ancient genomes from an ever wider
ple, although height is governed by hundreds of genes range of domesticated animal species. The data from
in humans125, only four loci explain over 80% of the var- these studies will not only enable more accurate recon-
iance in size in horses126. This example is the exception, struction of the history of genomic shifts associated with
however, and although tackling this genomic complexity the long-​term pattern of animal domestication but will
is challenging, new methods are being developed to trace also foster a sophisticated understanding of the processes
polygenic adaptation127. These efforts will also benefit that have led to the emergence of the modern world.
from the ongoing efforts of the Functional Annotation
Q11 of Animal Genomes (FAANG) Consortium, which aims
 Published online xx xx xxxx


1. Zeder, M. A. The domestication of animals. into the origins and spread of animal husbandry. 11. Pitulko, V. V. & Kasparov, A. K. Archaeological dogs
J. Anthropol. Res. 68, 161–190 (2012). J. Archaeol. Sci. 38, 538–545 (2011). from the Early Holocene Zhokhov site in the Eastern
2. Vigne, J.-​D. The origins of animal domestication 7. Vigne, J.-D. in The Neolithic Demographic Transition Siberian Arctic. J. Archaeol. Sci. Rep. 13, 491–515
and husbandry: a major change in the history of and its Consequences (eds Bocquet-​Appel, J.-​P. & (2017).
humanity and the biosphere. C. R. Biol. 334, Bar-​Yosef, O.) 179–205 (Springer, 2008). 12. Olsen, S. L. Early horse domestication: weighing the
171–181 (2011). 8. Ervynck, A., Dobney, K., Hongo, H. & Meadow, R. evidence. BAR. Int. Ser. 1560, 81 (2006).
3. Larson, G. et al. Rethinking dog domestication by Born free? New evidence for the status of ‘Sus scrofa’ 13. Larson, G. et al. Worldwide phylogeography of wild
integrating genetics, archeology, and biogeography. at Neolithic Çayönü Tepesi (Southeastern Anatolia, boar reveals multiple centers of pig domestication.
Proc. Natl Acad. Sci. USA 109, 8878–8883 (2012). Turkey). Paléorient 27, 47–73 (2001). Science 307, 1618–1621 (2005).
4. Darwin, C. The Variation of Animals and Plants Under 9. Payne, S. Kill-​off patterns in sheep and goats: the 14. Luikart, G. et al. Multiple maternal origins and
Domestication (John Murray, 1868). mandibles from Aşvan kale. Anatol. Stud. 23, weak phylogeographic structure in domestic
5. McHugo, G. P., Dover, M. J. & MacHugh, D. E. 281–303 (1973). goats. Proc. Natl Acad. Sci. USA 98, 5927–5932
Unlocking the origins and biology of domestic animals 10. Balasse, M. et al. Wild, domestic and feral? (2001).
using ancient DNA and paleogenomics. BMC Biol. 17, Investigating the status of suids in the Romanian 15. Naderi, S. et al. The goat domestication process
98 (2019). Gumelniţa (5th mil. cal BC) with biogeochemistry and inferred from large-​scale mitochondrial DNA analysis
6. Conolly, J. et al. Meta-​analysis of zooarchaeological geometric morphometrics. J. Anthropol. Archaeol. 42, of wild and domestic individuals. Proc. Natl Acad. Sci.
data from SW Asia and SE Europe provides insight 27–36 (2016). USA 105, 17659–17664 (2008).

www.nature.com/nrg
 Reviews

16. Pedrosa, S. et al. Evidence of three maternal lineages in European and Levantine cattle populations. replacement in European dogs during the Neolithic
Near Eastern sheep supporting multiple domestication A major shift in genomes occurred ~4,000 years Age as predicted by Frantz et al. (Science, 2016),
events. Proc. Biol. Sci. 272, 2211–2217 (2005). ago, with widespread introgression of zebu contradicting the hypothesis that dogs were
17. Vilà, C. et al. Widespread origins of domestic horse ancestry from the East, possibly linked to domesticated more than once.

Q14
Q13 lineages. Science 291, 474–477 (2001). multi-century drought. 58.

Ní Leathlobhair, M. et al. The evolutionary history of

18. Eriksson, J. et al. Identification of the yellow skin gene 42. Daly, K. G. et al. Ancient goat genomes reveal mosaic dogs in the Americas. Science 361, 81–85 (2018).
reveals a hybrid origin of the domestic chicken. PLoS domestication in the Fertile Crescent. Science 361, This study shows that American dogs are the
Genet. 4, e1000010 (2008). 85–88 (2018). descendants of dogs that dispersed with humans
19. Wright, D. Article commentary: the genetic This study is the first genome-​wide investigation of from Siberia into the Americas. Analyses of these
architecture of domestication in animals. Bioinform. ancient variation in a Fertile Crescent domesticate, genomes revealed that this ancient population of
Biol. Insights 9S4, BBI.S28902 (2015). the goat, and clearly shows the mosaic nature of dogs almost completely disappeared after the
20. Shannon, L. M. et al. Genetic structure in village dogs domestication via Neolithic Iranian, Anatolian and arrival of Europeans and that the last remnant of
reveals a Central Asian domestication origin. Proc. Levantine goat populations and their asymmetric this lineage is the genome of the canine transmissible
Natl Acad. Sci. USA 112, 13639–13644 (2015). relationships to pre-​domestication wild genomes. venereal tumour, a contagious cancer clone.
21. Wang, G.-D. et al. Out of southern East Asia: the The study also shows 8,000-​year-​old evidence for 59. Leonard, J. A. et al. Ancient DNA evidence for Old
natural history of domestic dogs across the world. selection of genes linked to pigmentation and other World origin of New World dogs. Science 298,
Cell Res. 26, 21–33 (2016). traits. 1613–1616 (2002).
22. Pang, J.-F. et al. mtDNA data indicate a single origin 43. Frantz, L. A. F. et al. Ancient pigs reveal a near-​complete 60. Frantz, L. A. F. & Larson, G. in Hybrid Communities
for dogs south of Yangtze River, less than 16,300 genomic turnover following their introduction to Europe. (eds Stépanoff, C. & Vigne, J.-​D.) 23–37 (Routledge,
years ago, from numerous wolves. Mol. Biol. Evol. 26, Proc. Natl Acad. Sci. USA 116, 17231–17238 (2019). 2018).
2849–2864 (2009). This paper reveals that the Near Eastern ancestry in 61. Larson, G. & Burger, J. A population genetics view of
23. Higuchi, R., Bowman, B., Freiberger, M., Ryder, O. A. the genomes of European domestic pigs, which is animal domestication. Trends Genet. 29, 197–205
& Wilson, A. C. DNA sequences from the quagga, an associated with the first domestic pigs that were (2013).
extinct member of the horse family. Nature 312, introduced into Europe around 8,000 years ago 62. Park, S. D. E. et al. Genome sequencing of the extinct
282–284 (1984). from the Near East, disappeared over 3,000 years Eurasian wild aurochs, Bos primigenius, illuminates
24. Pääbo, S. Ancient DNA: extraction, characterization, ago as a result of interbreeding with local wild boars. the phylogeography and evolution of cattle. Genome
molecular cloning, and enzymatic amplification. 44. Colledge, S., Conolly, J., Dobney, K., Manning, K. & Biol. 16, 234 (2015).
Proc. Natl Acad. Sci. USA 86, 1939–1943 (1989). Shennan, S. Origins and Spread of Domestic Animals This paper shows that the first ancient cattle
25. Hagelberg, E. et al. Analysis of ancient bone DNA: in Southwest Asia and Europe (Left Coast, 2013). genome sequenced was that of a Mesolithic British
techniques and applications. Philos. Trans. R. Soc. 45. Frantz, L. et al. The evolution of Suidae. Annu. Rev. auroch. Although there had been almost no trace
Lond. B Biol. Sci. 333, 399–407 (1991). Anim. Biosci. 4, 61–85 (2016). found of European auroch mtDNA in modern
26. Römpler, H. et al. The rise and fall of the 46. Helmer, D., Gourichon, L., Monchot, H., Peters, J. & cattle, these genome data show that detectable,
chemoattractant receptor GPR33. J. Biol. Chem. 280, Segui, M. S. in The First Steps of Animal probably male-​mediated, wild introgression had
31068–31075 (2005). Domestication (eds D. Vigne, J., Peters, J. & occurred within Europe.
27. Hofreiter, M., Serre, D., Poinar, H. N., Kuch, M. & Helmer, D.). 86–95 (Oxbow Books, 2005). 63. Frantz, L. A. F. et al. Evidence of long-​term gene flow
Pääbo, S. Ancient DNA. Nat. Rev. Genet. 2, 353–359 47. Hongo, H., Pearson, J., Öksüz, B. & Ilgezdi, G. The and selection during domestication from analyses of
(2001). process of ungulate domestication at Cayönü, Eurasian wild and domestic pig genomes. Nat. Genet.
28. Gilbert, M. T. P., Bandelt, H.-J., Hofreiter, M. & Southeastern Turkey: a multidisciplinary approach 47, 1141–1148 (2015).
Barnes, I. Assessing ancient DNA studies. Trends Ecol. focusing on Bos sp. and Cervus elaphus. This paper reveals that extensive gene flow
Evol. 20, 541–544 (2005). Anthropozoologica 44, 63–78 (2009). between wild and domestic populations took place
29. Goodwin, S., McPherson, J. D. & McCombie, W. R. 48. Perri, A. A wolf in dog’s clothing: initial dog during the evolutionary history of pigs and
Coming of age: ten years of next-​generation sequencing domestication and Pleistocene wolf variation. hypothesizes that an ‘island of domestication’
technologies. Nat. Rev. Genet. 17, 333–351 (2016). J. Archaeol. Sci. 68, 1–4 (2016). exists in the genome of domestic animals.
30. Rasmussen, M. et al. Ancient human genome 49. Drake, A. G., Coquerelle, M. & Colombeau, G. 3D 64. Skoglund, P., Ersmark, E., Palkopoulou, E. & Dalén, L.
sequence of an extinct Palaeo-​Eskimo. Nature 463, morphometric analysis of fossil canid skulls Ancient wolf genome reveals an early divergence of
757–762 (2010). contradicts the suggested domestication of dogs domestic dog ancestors and admixture into high-​
31. Green, R. E. et al. A draft sequence of the Neandertal during the late Paleolithic. Sci. Rep. 5, 8299 (2015). latitude breeds. Curr. Biol. 25, 1515–1519 (2015).
genome. Science 328, 710–722 (2010). 50. Outram, A. K. et al. The earliest horse harnessing and This paper reports the first genome-​wide data
32. Orlando, L., Gilbert, M. T. P. & Willerslev, E. milking. Science 323, 1332–1335 (2009). retrieved from an ancient canid and shows that
Reconstructing ancient genomes and epigenomes. 51. Frantz, L. A. F. et al. Genomic and archaeological there was gene flow between arctic dogs and a now
Nat. Rev. Genet. 16, 395–408 (2015). evidence suggest a dual origin of domestic dogs. extinct population of Siberian wolf.
This paper presents a comprehensive review of the Science 352, 1228–1231 (2016). 65. Marshall, F. B., Dobney, K., Denham, T. & Capriles, J. M.
techniques for extraction and sequencing of This paper presents the first analysis of an ancient Evaluating the roles of directed breeding and gene
ancient DNA. dog genome. Combined with a comprehensive flow in animal domestication. Proc. Natl Acad. Sci.
33. Haak, W. et al. Massive migration from the steppe was archaeological survey, the authors argue for a dual USA 111, 6153–6158 (2014).
a source for Indo-​European languages in Europe. origin of domestic dogs. 66. Schubert, M. et al. Prehistoric genomes reveal the
Nature 522, 207–211 (2015). 52. Freedman, A. H. et al. Genome sequencing highlights genetic foundation and cost of horse domestication.
34. Albrechtsen, A., Nielsen, F. C. & Nielsen, R. the dynamic early history of dogs. PLoS Genet. 10, Proc. Natl Acad. Sci. USA 111, E5661–E5669 (2014).
Ascertainment biases in SNP chips affect measures e1004016 (2014). 67. Turner, T. L., Hahn, M. W. & Nuzhdin, S. V. Genomic
of population divergence. Mol. Biol. Evol. 27, 53. Warmuth, V. et al. Reconstructing the origin and islands of speciation in Anopheles gambiae. PLoS Biol.
2534–2547 (2010). spread of horse domestication in the Eurasian steppe. 3, e285 (2005).
35. Boessenkool, S. et al. Combining bleach and mild Proc. Natl Acad. Sci. USA 109, 8202–8206 (2012). 68. Briggs, W. H., McMullen, M. D., Gaut, B. S. &
predigestion improves ancient DNA recovery from 54. The Bovine Genome Sequencing and Analysis Doebley, J. Linkage mapping of domestication loci in a
bones. Mol. Ecol. Resour. 17, 742–751 (2017). Consortium. et al. The genome sequence of taurine large maize–teosinte backcross resource. Genetics
36. Gansauge, M.-T. & Meyer, M. Single-​stranded DNA cattle: a window to ruminant biology and evolution. 177, 1915–1928 (2007).
library preparation for the sequencing of ancient or Science 324, 522–528 (2009). 69. Trut, L. Early canid domestication: the Farm-​Fox
damaged DNA. Nat. Protoc. 8, 737–748 (2013). 55. Fages, A. et al. Tracking five millennia of horse Experiment Foxes bred for tamability in a 40-​year
37. Gansauge, M.-T. et al. Single-​stranded DNA library management with extensive ancient genome time experiment exhibit remarkable transformations that
preparation from highly degraded DNA using T4 DNA series. Cell 177, 1419–1435.e31 (2019). suggest an interplay between behavioral genetics and
ligase. Nucleic Acids Res. 45, e79 (2017). This paper represents a milestone in ancient development. Am. Sci. 87, 160–169 (1999).
38. Briggs, A. W. et al. Removal of deaminated cytosines animal genomics, with the sequencing of over 120 70. Lord, K. A., Larson, G., Coppinger, R. P. &
and detection of in vivo methylation in ancient DNA. ancient horse genomes. The authors draw multiple Karlsson, E. K. The history of farm foxes undermines
Nucleic Acids Res. 38, e87 (2010). key conclusions related to the genetic make-​up of the animal domestication syndrome. Trends Ecol. Evol.
39. Rohland, N., Harney, E., Mallick, S., Nordenfelt, S. & modern domestic horses and the fitness cost of 35, 125–136 (2020).
Reich, D. Partial uracil–DNA-​glycosylase treatment for modern artificial selection. 71. Librado, P. et al. Ancient genomic changes associated
screening of ancient DNA. Philos. Trans. R. Soc. Lond. 56. Gaunitz, C. et al. Ancient genomes revisit the ancestry with domestication of the horse. Science 356,
B Biol. Sci. 370, 20130624 (2015). of domestic and Przewalski’s horses. Science 360, 442–445 (2017).
40. Gamba, C. et al. Genome flux and stasis in a five 111–114 (2018). 72. Pendleton, A. L. et al. Comparison of village dog and
millennium transect of European prehistory. This study uses ancient horse genomes to show that wolf genomes highlights the role of the neural crest in
Nat. Commun. 5, 5257 (2014). the Botai horses, which are thought to be the first dog domestication. BMC Biol. 16, 64 (2018).
This paper identifies the petrous temporal bone as domestic horses, are not related to modern horses 73. Marsden, C. D. et al. Bottlenecks and selective sweeps
the best reservoir of preserved ancient DNA in but instead form a lineage that is now almost during domestication have increased deleterious
human remains — a finding that extends to extinct. The sole surviving representative of this genetic variation in dogs. Proc. Natl Acad. Sci. USA
archaeological specimens of domesticated animals lineage is the modern Przewalski’s horse, which was 113, 152–157 (2016).
and that enables the frequent recovery of whole-​ thought to be the last wild horse but is in fact the 74. MacLeod, I. M., Larkin, D. M., Lewin, H. A., Hayes, B. J.
genome sequences using shotgun sequencing. descendant of an ancient domesticated population. & Goddard, M. E. Inferring demography from runs
41. Verdugo, M. P. et al. Ancient cattle genomics, origins, 57. Botigué, L. R. et al. Ancient European dog genomes of homozygosity in whole-​genome sequence, with
and rapid turnover in the Fertile Crescent. Science reveal continuity since the Early Neolithic. Nat. correction for sequence errors. Mol. Biol. Evol. 30,
365, 173–176 (2019). Commun. 8, 16082 (2017). 2209–2223 (2013).
This paper presents genome data from ancient This paper describes the analysis of two newly 75. Carneiro, M. et al. Rabbit genome analysis reveals
Near Eastern cattle and aurochs showing sequenced Neolithic dog genomes from Germany, a polygenic basis for phenotypic change during
introgression of local aurochs into both early which revealed that there was no population domestication. Science 345, 1074–1079 (2014).

Nature Reviews | Genetics

Reviews

76. Allaby, R. G., Ware, R. L. & Kistler, L. A re-​evaluation 103. Ludwig, A. et al. Coat color variation at the beginning 131. Teasdale, M. D. et al. Paging through history:
of the domestication bottleneck from archaeogenomic of horse domestication. Science 324, 485 (2009). parchment as a reservoir of ancient DNA for next
evidence. Evol. Appl. 12, 29–37 (2019). 104. Linderholm, A. & Larson, G. The role of humans generation sequencing. Philos. Trans. R. Soc. Lond. B
77. Bosse, M. et al. Regions of homozygosity in the in facilitating and sustaining coat colour variation in Biol. Sci. 370, 20130379 (2015).
porcine genome: consequence of demography and the domestic animals. Semin. Cell Dev. Biol. 24, 587–593 132. O’Sullivan, N. J. et al. A whole mitochondria analysis
recombination landscape. PLoS Genet. 8, e1003100 (2013). of the Tyrolean Iceman’s leather provides insights into
(2012). 105. Pavlidis, P., Jensen, J. D. & Stephan, W. Searching for the animal sources of Copper Age clothing. Sci. Rep.
78. Bollongino, R. et al. Modern taurine cattle descended footprints of positive selection in whole-​genome SNP 6, 31279 (2016).
from small number of near-​eastern founders. Mol. data from nonequilibrium populations. Genetics 185, 133. Bro-​Jørgensen, M. H. et al. Ancient DNA analysis of
Biol. Evol. 29, 2101–2104 (2012). 907–922 (2010). Scandinavian medieval drinking horns and the horn of
79. Murray, C., Huerta-​Sanchez, E., Casey, F. & 106. Schraiber, J. G., Evans, S. N. & Slatkin, M. Bayesian the last aurochs bull. J. Archaeol. Sci. 99, 47–54
Bradley, D. G. Cattle demographic history modelled inference of natural selection from allele frequency (2018).
from autosomal sequence variation. Philos. Trans. R. time series. Genetics 203, 493–511 (2016). 134. Zeder, M. A. in Harlan II: Biodiversity in Agriculture:
Soc. Lond. B Biol. Sci. 365, 2531–2539 (2010). 107. Malaspinas, A.-S. Methods to characterize selective Domestication, Evolution and Sustainability (eds
80. Kristiansen, K. The Rise of Bronze Age Society: sweeps using time serial samples: an ancient DNA Damania, A. & Gepts, P.) 227–229 (Univ. California,
Travels, Transmissions and Transformations. perspective. Mol. Ecol. 25, 24–41 (2016). 2011).
(Cambridge University Press, 2005). 108. Barrios-​Garcia, M. N. & Ballari, S. A. Impact of wild 135. Hanotte, O. et al. African pastoralism: genetic imprints
81. Allentoft, M. E. et al. Population genomics of Bronze boar (Sus scrofa) in its introduced and native range: of origins and migrations. Science 296, 336–339
Age Eurasia. Nature 522, 167–172 (2015). a review. Biol. Invasions 14, 2283–2300 (2012). (2002).
82. Damgaard, P. et al. 137 ancient human genomes from 109. Boyd, J. M., Doney, J. M., Gunn, R. G. & Jewell, P. A. 136. Freeman, A. R., Bradley, D. G., Nagda, S., Gibson, J. P.
across the Eurasian steppes. Nature 557, 369–374 in Proc. Zool. Soc. London Vol. 142, 129–164 (Wiley & Hanotte, O. Combination of multiple microsatellite
(2018). Online Library, 1964). data sets to investigate genetic diversity and admixture
83. Jeong, C. et al. Bronze Age population dynamics and 110. Ariefiandy, A. et al. Temporal and spatial dynamics of of domestic cattle. Anim. Genet. 37, 1–9 (2006).
the rise of dairy pastoralism on the eastern Eurasian insular Rusa deer and wild pig populations in Komodo 137. Rubin, C.-J. et al. Whole-​genome resequencing reveals
steppe. Proc. Natl Acad. Sci. USA 115, National Park. J. Mammal. 97, 1652–1662 (2016). loci under selection during chicken domestication.
E11248–E11255 (2018). 111. Vigne, J.-D. et al. Pre-​Neolithic wild boar management Nature 464, 587–591 (2010).
84. Hansen, P. J. Physiological and cellular adaptations of and introduction to Cyprus more than 11,400 years 138. Karlsson, A.-C. et al. The effect of a mutation in the
zebu cattle to thermal stress. Anim. Reprod. Sci. ago. Proc. Natl Acad. Sci. USA 106, 16135–16138 thyroid stimulating hormone receptor (TSHR) on
82-83, 349–360 (2004). (2009). development, behaviour and TH levels in domesticated
85. Matthews, R. Zebu: harbingers of doom in Bronze Age 112. Heinsohn, T. Animal translocation: long-​term human chickens. PLoS One 10, e0129040 (2015).
western Asia? Antiquity 76, 438–446 (2002). influences on the vertebrate zoogeography of 139. Fang, M., Larson, G., Ribeiro, H. S., Li, N. &
86. Ollivier, M. et al. Dogs accompanied humans during Australasia (natural dispersal versus ethnophoresy). Andersson, L. Contrasting mode of evolution at a coat
the Neolithic expansion into Europe. Biol. Lett. 14, Aust. Zool. 32, 351–376 (2003). color locus in wild and domestic pigs. PLoS Genet. 5,
20180286 (2018). 113. Taberlet, P. et al. Are cattle, sheep, and goats e1000341 (2009).
87. Ottoni, C. et al. Pig domestication and human-​mediated endangered species? Mol. Ecol. 17, 275–284 (2008). 140. Andersson, L. S. et al. Mutations in DMRT3 affect
dispersal in western Eurasia revealed through ancient 114. Taberlet, P., Coissac, E., Pansu, J. & Pompanon, F. locomotion in horses and spinal circuit function in
DNA and geometric morphometrics. Mol. Biol. Evol. Conservation genetics of cattle, sheep, and goats. mice. Nature 488, 642–646 (2012).
30, 824–832 (2013). C. R. Biol. 334, 247–254 (2011). 141. Wutke, S. et al. The origin of ambling horses. Curr.
88. White, S. From globalized pig breeds to capitalist pigs: 115. Heckenberger, M. J., Russell, J. C., Toney, J. R. & Biol. 26, R697–R699 (2016).
a study in animal cultures and evolutionary history. Schmidt, M. J. The legacy of cultural landscapes in the 142. Li, H. & Durbin, R. Fast and accurate short read
Environ. Hist. Durh. N. C. 16, 94–120 (2011). Brazilian Amazon: implications for biodiversity. Philos. alignment with Burrows–Wheeler transform.
89. Bosse, M. et al. Genomic analysis reveals selection for Trans. R. Soc. Lond. B Biol. Sci. 362, 197–208 (2007). Bioinformatics 25, 1754–1760 (2009).
Asian genes in European pigs following human-​ 116. Ellis, E. C. et al. Used planet: a global history. Proc. 143. Schubert, M. et al. Improving ancient DNA read
mediated introgression. Nat. Commun. 5, 4392 (2014). Natl Acad. Sci. USA 110, 7978–7985 (2013). mapping against modern reference genomes.
90. Bosse, M. et al. Artificial selection on introduced 117. Piperno, D. R., McMichael, C. & Bush, M. B. Amazonia BMC Genomics 13, 178 (2012).
Asian haplotypes shaped the genetic architecture in and the Anthropocene: what was the spatial extent 144. Rajaraman, A., Tannier, E. & Chauve, C. FPSAC: fast
European commercial pigs. Proc. Biol. Sci. 282, and intensity of human landscape modification in the phylogenetic scaffolding of ancient contigs.
20152019 (2015). Amazon basin at the end of prehistory? Holocene 25, Bioinformatics 29, 2987–2994 (2013).
91. Shearin, A. L. & Ostrander, E. A. Canine morphology: 1588–1597 (2015). 145. Seitz, A. & Nieselt, K. Improving ancient DNA genome
hunting for genes and tracking mutations. PLoS Biol. 118. Plug, I. Aspects of life in the Kruger National Park assembly. PeerJ 5, e3126 (2017).
8, e1000310 (2010). during the early iron age. Goodwin Series 6, 62–68 146. Patterson, N. et al. Ancient admixture in human
92. Boyko, A. R. The domestic dog: man’s best friend in (1989). history. Genetics 192, 1065–1093 (2012).
the genomic era. Genome Biol. 12, 216 (2011). 119. Spyrou, M. A., Bos, K. I., Herbig, A. & Krause, J. 147. Patterson, N., Price, A. L. & Reich, D. Population
93. Bosse, M. et al. Untangling the hybrid nature of Ancient pathogen genomics as an emerging tool for structure and eigenanalysis. PLoS Genet. 2, e190
modern pig genomes: a mosaic derived from infectious disease research. Nat. Rev. Genet. 20, (2006).
biogeographically distinct and highly divergent Sus 323–340 (2019). 148. Alexander, D. H., Novembre, J. & Lange, K. Fast
scrofa populations. Mol. Ecol. 23, 4089–4102 (2014). 120. Roman-​Binois, A. L’archéologie des épizooties: mise en model-​based estimation of ancestry in unrelated
94. Orlando, L. & Librado, P. Origin and evolution of évidence et diagnostic des crises de mortalité chez les individuals. Genome Res. 19, 1655–1664 (2009).
deleterious mutations in horses. Genes 10, E649 animaux d’élevage, du Néolithique à Pasteur. 149. Durand, E. Y., Patterson, N., Reich, D. & Slatkin, M.
(2019). (Université Panthéon-​Sorbonne-​Paris I, 2017). Testing for ancient admixture between closely related
95. Oltenacu, P. A. & Algers, B. Selection for increased 121. Boodhoo, N., Gurung, A., Sharif, S. & Behboudi, S. populations. Mol. Biol. Evol. 28, 2239–2252 (2011).
production and the welfare of dairy cows: are new Marek’s disease in chickens: a review with focus on 150. Pickrell, J. K. & Pritchard, J. K. Inference of population
breeding goals needed? Ambio 34, 311–315 (2005). immunology. Vet. Res. 47, 119 (2016). splits and mixtures from genome-​wide allele frequency
96. Charlier, C. et al. NGS-​based reverse genetic screen for 122. Witter, R. L. The changing landscape of Marek’s data. PLoS Genet. 8, e1002967 (2012).
common embryonic lethal mutations compromising disease. Avian Pathol. 27, S46–S53 (1998).
fertility in livestock. Genome Res. 26, 1333–1341 123. Bellone, R. R. et al. Evidence for a retroviral insertion Acknowledgements 
(2016). in TRPM1 as the cause of congenital stationary night L.A.F.F. and G.L. were supported by a European Research Q12

97. Derks, M. F. L. et al. Loss of function mutations in blindness and leopard complex spotting in the horse. Council (ERC) grant (ERC-2013–396 StG-337574-​UNDEAD)
essential genes cause embryonic lethality in pigs. PLoS One 8, e78280 (2013). and/or Natural Environmental Research Council grants
PLoS Genet. 15, e1008055 (2019). 124. FAO. Domestic animal diversity information system. (NE/K005243/1, NE/K003259/1, NE/S00078X/1, NE/
98. Ellegren, H. The different levels of genetic diversity in DAD-​IS http://dad.fao.org/ (2017). S007067/1). L.A.F.F. was also supported by a Wellcome Trust
sex chromosomes and autosomes. Trends Genet. 25, 125. Wood, A. R. et al. Defining the role of common grant (210119/Z/18/Z). L.O. was supported by the ERC under
278–284 (2009). variation in the genomic and biological architecture the European Union’s Horizon 2020 research and innovation
99. Peters, J., Arbuckle, B. S. & Pöllath, N. in The Neolithic of adult human height. Nat. Genet. 46, 1173–1186 programme (grant agreement No 681605). D.G.B. was
in Turkey Vol. 6 (eds Özdogan, M., Baflgelen, N. & (2014). supported by ERC Investigator grant 295729-​CodeX.
Kuniholm, P.). 1–65 (Archaeology and Art, 2012). 126. Makvandi-​Nejad, S. et al. Four loci explain 83% of size
100. Wallner, B. et al. Y chromosome uncovers the recent variation in the horse. PLoS One 7, e39929 (2012). Author contributions
oriental origin of modern stallions. Curr. Biol. 27, 127. Racimo, F., Sankararaman, S., Nielsen, R. & The authors contributed equally to all aspects of the article.
2029–2035.e5 (2017). Huerta-​Sánchez, E. Evidence for archaic adaptive
101. Girdland Flink, L. et al. Establishing the validity of introgression in humans. Nat. Rev. Genet. 16, Competing interests
domestication genes using DNA from ancient 359–371 (2015). The authors declare no competing interests.
chickens. Proc. Natl Acad. Sci. USA 111, 6184–6189 128. Andersson, L. et al. Coordinated international action
(2014). to accelerate genome-​to-​phenome with FAANG, the Peer review information
This study uses ancient DNA to show that Functional Annotation of Animal Genomes Project. Nature Reviews Genetics thanks S. Olsen, P. Ajmone-​Marsan
mutations that are fixed in modern populations Genome Biol. 16, 57 (2015). and the other, anonymous, reviewer(s) for their contribution
and which were thought to be important during 129. Warinner, C. et al. Pathogens and host immunity in the to the peer review of this work.
chicken domestication were in fact only the target ancient human oral cavity. Nat. Genet. 46, 336–344
of recent artificial selection. (2014). Publisher’s note
102. Loog, L. et al. Inferring allele frequency trajectories 130. Schubert, M. et al. Zonkey: a simple, accurate and Springer Nature remains neutral with regard to jurisdictional
from ancient DNA indicates that selection on a chicken sensitive pipeline to genetically identify equine claims in published maps and institutional affiliations.
gene coincided with changes in medieval husbandry F1-​hybrids in archaeological assemblages. J. Archaeol.
practices. Mol. Biol. Evol. 34, 1981–1990 (2017). Sci. 78, 147–157 (2017). © Springer Nature Limited 2020

www.nature.com/nrg
 QUERY FORM
Nature Reviews Genetics
Manuscript ID 225

Author Laurent A. F. Frantz

AUTHOR:

The following queries have arisen during the editing of your manuscript. Please answer by making the requisite corrections directly
in the e.proofing tool rather than marking them up on the PDF. This will ensure that your corrections are incorporated accurately
and that your paper is published as quickly as possible.

Query No. Nature of Query

Q1: AU: Reference [8] was a duplicate of [6] and hence the repeated version has been deleted. Please check.

Q2: Please check your article carefully , coordinate with any co-​authors and enter all final edits clearly in the eproof,
remembering to save frequently. Once corrections are submitted, we cannot routinely make further changes to
the article.

Q3: Note that the eproof should be amended in only one browser window at any one time; otherwise changes will be
overwritten.

Q4: Author surnames have been highlighted. Please check these carefully and adjust if the first name or surname is
marked up incorrectly. Note that changes here will affect indexing of your article in public repositories such as
PubMed. Also, carefully check the spelling and numbering of all author names and affiliations, and the corre-
sponding email address(es).

Q5: AU: Please state in Fig. 2 caption what yellow represents

Q6: Au: Edit in sentence starting “The latter is…” to distinguish from Iberia centered on present-​day Eastern Georgia
OK?

Q7: AU: Please confirm that edits to the sentence starting “For example, a major …” are OK

Q8: AU: Please define f4 in Fig. 4 caption

Q9: Au: Edit in the sentence starting “However, the dynamics …” OK? As we avoid use of the word ‘Eskimo’.

Q10: AU: Please confirm that edits to the sentence starting “In cattle, for …” are OK

Q11: AU: Please confirm definition “Functional Annotation of Animal Genomes (FAANG)”

Q12: AU: Please confirm that edits to the sentence starting “L.A.F.F. and G.L. were …” are OK

Q13: AU: Please confirm that edits to the blurbs for references 41, 42 and 62 are OK

Q14: AU: Please confirm “Frantz et al. (Science, 2016)” refers to the correct study (ref. 51), or provide revisions

Q15: Please keep the heading to a single line

Springer Nature