Skip to main content
NCBI home page
Journal of Heredity logoLink to Journal of Heredity
. 2022 Jan 25;113(2):160–170. doi: 10.1093/jhered/esac002

Genetic Origins of the Two Canis lupus familiaris (Dog) Freight Dog Populations

Muhammad Basil Ali 1,2,✉, Dayna L Dreger 3, Reuben M Buckley 4, Shahid Mansoor 5,6, Qaiser M Khan 7,8, Elaine A Ostrander 9,✉
Editor: Bridgett vonHoldt
PMCID: PMC9113510  PMID: 35575082

Abstract

Despite periodic drops in popularity, Arctic sled dogs continue to play a vital role in northern societies, providing both freight transit and recreational race activities. In this study, we selected the Mackenzie River Husky, a freight dog of complex history, and the Chinook, an American Kennel Club recognized freight dog breed whose heritage reportedly overlaps that of the MKRH, for detailed population analysis. We tested each to determine their component breeds and used admixture analysis to ascertain their population structure. We utilized haplotype analysis to identify genomic regions shared between each population and their founding breeds. Our data show that the Alaskan Malamutes and modern Greenland sled dog contributed to both populations, but there are also unexpected contributions from the German Shepherd dog and Collie. We used haplotype analysis to identify genomic regions nearing fixation in population type and identify provocative genes in each region. Finally, in response to recent reports regarding the importance of dietary lipid genes in Arctic dogs, we analyzed 8 such genes in a targeted analysis observing signatures of selection in both populations at the MLXIPL gene loci. These data highlight the genetic routes that breeds of similar function have taken toward their occupation as successful sled dogs.

Keywords: Arctic, canine, sled dog, haplotype

Since their earliest days, domestic dogs have been central to human survival, serving as guardians, herders, and protectors, as well as many other occupations. In the Arctic, they fill a unique niche (Serpell 1995). The use of dogs for pulling sledges dates to at least 9500 years before present (BP) in Siberia (Pitulko and Kasparav 2017), though the use of sled dogs in North America is relatively recent, dating to approximately 1000 BP (Morey and Aaris-Sorensen 2002; Ameen et al. 2019). At that time, the eastward expansion of the Inuit people of the Thule culture from Alaska to Eastern Canada and Greenland relied heavily on the transportation advantages provided by sled dogs (Ford 1959). Reliance of dog-powered conveyance persists in these regions to this day (Wendt 1999).

One interesting example of a modern functional freight dog is the Mackenzie River Husky (MKRH) (http://www.sleddogcentral.com/mackenzies2.htm), a “catch all” name that is used to describe several populations of similar appearing Arctic freight hauling dogs. The American Kennel Club does not recognize the MKRH, nor is there a club with defined standards, as MKRH were historically bred by different kennels seeking different attributes. The “breed,” as we refer to it in this paper, has many of the characteristics of established breeds like the Alaskan Malamute (AMAL) or Siberian Husky (HUSK), both well-established sled dog populations. MKRH are heavy (70–120 lbs), but with a lean build and long legs (25–32 inches) (Figure 1). As they are bred for freight work, although used occasionally for sled racing, they share many climate-specific athletic attributes, such as superior pulling strength and high endurance in cold weather with limited food sources.

Figure 1.

Figure 1.

Open in a new tab

Pedigree of MKRH dogs with DNA samples and photos. (A) Arthur T. Walden with his dog, Chinook, for whom the Chinook breed is named, circa 1922 (photograph in the public domain). (B) A modern Chinook dog (photograph courtesy of M. Mandt). (C) A modern Mackenzie River Husky dog (photograph adapted from the Creative Commons, https://commons.wikimedia.org/wiki/File:MackenzieRiverHusky.jpg). (D) Pedigree of the 21 MKRH dogs used in these analyses. Colored shapes (purple and green) are dogs for which DNA samples are available. Purple identifies the least related individuals of the sample set, used in admixture analysis. Those in gray are not available for sampling.

The MKRH history is complex, dating to the 1800s (https://www.sleddogcentral.com/mackenzies2.htm), with purported crosses to several breeds including the St. Bernard (STBD), Newfoundland (NEWF), and mastiff (MAST). Their development as a population has included at least a dozen name changes and formation of many subpopulations, but they have maintained a consistent niche of working dogs who haul freight. The MKRH decreased in popularity starting in the 1960s as snowmobiles and other means of transportation came into widespread use. Use of the dogs in racing, which sometimes happened, favored the development of smaller fast dogs (http://www.sleddogcentral.com/mackenzies2.htm), further decreasing the population, as did concern of rabies infections. Today, nearly all modern MKRH have ties to a single kennel in Fairbanks, AK, named Northern Quest Kennels, which has overseen the continuance of the MKRH for over 40 years. Given the various forces that have shaped MKRH dogs from that kennel, we were interested in developing a population genetics understanding of the MKRH Northern Quest Kennel Dogs as they exist today. While the story is apt to be complex, as all the MKRH studied will, of necessity, be from the same kennel, thus limiting extrapolation of our results, this study will provide us with insights as to the early history of the breed, as well as the ability to compare the breed to other similarly shaped sled dog breeds such as the Chinook (COOK).

The COOK breed is used today for both freight transfer and sled racing. Developed in the early 1900s in New Hampshire, the COOK is hypothesized to be a mixture of HUSK, modern Greenland sledge dog (GREE), Belgian shepherd (BELS), German shepherd dog (GSD), and mastiff (MAST) (https://chinook.org/history/). However, unlike the MKRH, the COOK breed has been recognized by the AKC since 2013 (American Kennel Club 2017). Like the MKRH, the COOK breed has a small population size, with 2020 registrations placing the COOK as 186th of 195 breeds (https://www.akc.org/expert-advice/dog-breeds/the-most-popular-dog-breeds-of-2020/), with about 800–900 dogs alive today. COOK are a mid-sized dog, standing between 22 and 27 inches at the withers (shoulders), generally of a tawny color with a close, dense coat, dark eyes, and a slightly curving tail (http://images.akc.org/pdf/breeds/standards/Chinook.pdf) (Figure 1). The COOK thus serves as a useful comparison for understanding the evolution of populations with small membership and similar tasks, but distinctive histories. We used haplotype sharing and admixture analysis to identify the genetic composition of both populations, and the constituent breeds that comprise each. We show that each contains loci of interest that are close to fixation that are shared by founding parental breeds.

Materials and Methods

Samples and Genotyping

All sample collection protocols in this study were approved by the National Human Genome Research Institute (NHGRI) Animal Care and Use Committee at the National Institutes of Health (NIH) (Protocol GFS-05-1). Blood samples from 21 MKRH were provided by Northern Quest Kennels in Fairbanks, Alaska. While the overall number of MKRH alive today is unknown, it is estimated to be less than 100. DNA was isolated using a proteinase K-SDS chloroform phenol extraction protocol (Sambrook et al. 1989). A previously reported dataset of 1346 dogs genotyped at 150 112 single-nucleotide polymorphisms (SNPs) using the Canine HD Whole-Genome Genotyping BeadChip (173 662 total SNPs) from Illumina (San Diego, CA) was utilized (Parker et al. 2017). That dataset included 161 breeds, ranging from 2 to 19 dogs per breed, and 9 wild canids, which included 7 gray wolves and 3 golden jackals (Supplementary Table S1). Genotype data for 10 COOKs, produced on the same genotyping array, were obtained from published sources (Hayward et al. 2016). Based on available information, the COOKs are assumed to be unrelated to one another. Genotype filtering for the MKRH was performed using Illumina Genome Studio v.2 (Zhao et al. 2018). Samples with >90% SNP call rate, Gentrain scores of >0.4, and a minor allele frequency of >1% were retained, and the 2 datasets merged using PLINK v.1.9 (Purcell et al. 2007). The final combined dataset contains 1377 domestic dogs and 9 wild canids. Genotype calls were based on CanFam 3.1 reference genome assembly (Lindblad-Toh et al. 2005). After quality control pruning, a total of 150 118 SNPs remained.

Pedigree Analysis

The MKRH pedigree revealed that the samples provided were from highly related dogs (Figure 1). The dataset contained 2 sets of 3 siblings and 2 sets of 2 siblings. All reported 0–3 generation pedigree relationships were verified with the relationship matrix function of PLINK (Purcell et al. 2007). In addition, multiple dogs were bred redundantly in the pedigree. The 7 most unrelated dogs: 77448–77450, 77452, 77454–77456, who did not share any common parents, were used for admixture analysis. Since we do not know if the current pedigrees are open to cross breeding or are closed breeding pedigrees, we utilized only dogs whose parentage we could ascertain (Figure 1).

Haplotype Sharing, Principal Component Analysis, and Admixture

To ascertain breeds contributing to the MKRH and COOK, identity by descent (IBD) haplotype sharing analysis was done using BEAGLE v.4.1 (Browning and Browning 2015). Each MKRH and each COOK were compared with every dog that was not a member of its own breed. Sharing was calculated using a window size of 2100 SNPs and an overlapping value of 50 SNPs. Additionally, the same window size and overlapping values were used to calculate IBD haplotype sharing using each of the test breeds (MKRH and COOK), and each of the remaining 161 previously genotyped breeds mentioned above. A significance level of 95% was calculated by summing the haplotype sharing of all possible pairs of dogs from different breeds using the quantile function in R. Breed pairs with ≥95% sharing were considered to be foundational. Within-breed IBD values were obtained for each breed analyzed by calculating the mean IBD of each same-breed pair of dogs.

Principal component analysis (PCA) was performed with all the breeds identified through IBD haplotype sharing analysis using FlashPCA software (Abraham et al. 2017). The dataset, in addition to MKRH and COOK, was comprised of 16 breeds: 10 each of AMAL, GREE, HUSK, GSD, COLL, Australian shepherd (AUSS), Peruvian Inca orchid (INCA), Portuguese water dog (PTWD), Leonberger (LEON), Shetland sheepdog (SSHP), 5 Belgian Malinois (BMAL), and Xoloitzcuintli (XOLO), 3 Berger Picards (BPIC), and 2 cane Paratores (CPAT). The ancestry-based admixture estimation was calculated with the same dataset used for the PCA analysis, and utilized Admixture v.1.3.0 (Alexander and Lange 2011), with 2–15 adjusted cluster ancestry models. The optimal cluster model (K) was determined using the cross-validation error procedure as defined by Admixture. Results are shown for K = 2–13, with K = 8 being the optimal. At K = 13, the COOK and MKRH populations, along with dogs from 11 other breeds, split as expected along breed lines. Ancestry models were graphed using the Cluster Markov Packager Across K (CLUMPAK) online server (Kopelman et al. 2015).

Region of Homozygosity and Hierarchal Clustering of Haplotypes

To ascertain regions with excess homozygosity within the MKRH and COOK, and to identify regions that are close to fixation, which we define as those in common at 100% and in a subsequent analysis in 90% of the total population, we calculated runs of homozygosity (ROH). ROH were computed for each MKRH and COOK dog using --homozyg function in PLINK with a minimum window size of 10 kb, requiring at least 5 sequential SNPs, and allowing for 1 heterozygous SNP per window (Purcell et al. 2007). All ROH identified in each dog were compiled in separate input files and an intersection analysis was conducted using Bedtools v2.30.0 (Quinlan and Hall 2010) to identify regions that were common to 100% and 90% of the population of each breed.

Regions that passed the threshold of 100% or 90% underwent haplotype analysis. Haplotypes were identified in each individual MKRH and COOK, and each of a set of 5 putative founder breeds (AMAL, HUSK, GREE, GSD, and COLL), as determined by haplotype sharing and admixture analysis. Haplotypes for each ROH were analyzed using DnaSP v.5.10.1 software (Librado and Rozas 2009), allowing determination of the total number of haplotypes and the haplotype frequency for each breed.

Hierarchical clustering was conducted using the average distance method such that similar haplotypes were grouped together after pairwise comparison of all the haplotypes using a 90% average similarity threshold. Haplotype groups were based on the number of differences between each haplotype (hamming distance). The clustering of haplotypes for each region is represented as a phylogenetic dendrogram.

Nucleotide Diversity Analysis of Genes Involved in Dietary Metabolism

Because recent publications suggest a role for dietary genes in lipid processing in sled dogs (Sinding 2020), we selected 8 dietary lipid metabolism genes for an in-depth investigation: APOO, CERS5, FOXO1, FOXS1, MLXIPL, PDE10A, TM6SF2, and SLC25A40 (Rosenthal et al. 2013; Turkieh et al. 2014; Sinding 2020). Genotypes from chip array SNPs within the genomic region of each dietary gene, plus 2.5Mb of up- and down-stream flanking regions, were extracted from each MKRH, COOK, and 10 village dogs (VILL) as a control population. The control population was selected because VILL lack both human-directed selective pressure and breed structure (Shannon et al. 2015). We assessed nucleotide diversity (π) in windows of 200 kb using vcftools in each of the populations (MKRH, COOK, VILL). For each 200 kb window, we calculated the difference in π (Δ π) between control and target populations using the equation Δ π = π(control) – π(target). Positive Δ π values suggest loss of diversity in the target population (π(control) > π(target)), while negative Δ π values suggest loss of diversity in the control population (π(control) < π(target)). Windows with positive Δ π values in the empirical top 5% underwent further investigation.

Results

PCA, Admixture, and Haplotype Sharing

In this study, we explored the breed structure of 2 unique sled dog populations, the MKRH and the COOK, with a specific focus on identifying common foundational breeds and shared genomic history. Our goal was to ascertain the degree to which sled dogs of distinct developmental profiles, but similar occupations, shared genomic regions. To determine the component breeds that make up the MKRH and COOK populations, we performed IBD haplotype sharing analysis for each individual dog with an established comparative dataset (Parker et al. 2017) of 1346 dogs representing 161 breeds. The 21 MKRHs in this analysis are comprised of a large extended pedigree with a high degree of relatedness, potentially inflating the amount of haplotype sharing within the breed. Significant haplotype sharing was noted between MKRH and HUSK, GREE, AMAL, CPAT, BPIC, GSD, and COLL (Figure 2A). The levels of haplotype sharing between the MKRH with the CPAT and BPIC are lower than that of the GSD which, when considering known relationships between the breeds (Parker et al. 2017; Talenti et al. 2019), indicates that the CPAT and BPIC signals with the MKRH are likely present due to the shared ancestry of the breeds with the GSD.

Figure 2.

Figure 2.

Open in a new tab

IBD haplotype sharing boxplot of MKRH and COOK with 161 breeds. (A) The percent of MKRH and COOK dogs that show significant haplotype sharing with each of the potential founder breeds. (B and C) Breed-to-breed IBD haplotype sharing. Each vertical bar represents the amount of haplotype sharing with the sequentially listed breed and either (B) MKRH or (C) COOK. Colors differentiate previously defined clades (Parker et al. 2017). Breed-to-breed haplotype sharing is considered significant if the breed median is greater than the 95% dog-to-dog across-breed sharing (horizontal black line). Breed abbreviations of component breeds with significant haplotype sharing are listed.

By comparison, IBD haplotype sharing analysis with the COOK revealed significant commonalities with GREE, AMAL, PTWD, INCA, XOLO, CPAT, BPIC, GSD, BMAL, COLL, SSHP, AUSS, and LEON (Figure 2B). As with the MKRH, shared relatedness to the GSD likely accounts for the PTWD, INCA, XOLO, CPAT, BPIC, LEON, and BMAL signals. The SSHP and AUSS signals can likewise be attributed to the shared relationships with COLL. As such, the foundational history of the MKRH is made up predominantly of the HUSK, GREE, and AMAL sled dog breeds, and the GSD and COLL herding breeds. The COOK, likewise, has a foundational history tracing to the GREE and AMAL sled dog breeds and the GSD and COLL herding breeds. The difference between MKRH and COOK, therefore, is the inclusion of HUSK with the MKRH and a greater or more recent contribution of GSD and COLL to the COOK.

Within-breed mean IBD values were calculated to assess the degree of homogeneity within the MKRH and COOK breeds, relative to the 161 established breed populations in the comparative dataset (Supplementary Table S1). The average IBD value is 1.59 × 109 for the MKRH and 3.29 × 109 for the COOK. For comparison, the lowest breed average IBD obtained for the 161 additional breeds is 4.62 × 107 for the Mastino Abruzzese, and the highest value is 4.41 × 109 for the BPIC. The degree of within-breed haplotype sharing in the MKRH is comparable to the Vizsla (1.58 × 109) and Spinone Italiano (1.60 × 109), while the degree of haplotype sharing within the COOK breed is comparable to the Norwich Terrier (3.27 × 109) or Wire Fox Terrier (3.30 × 109).

To further explore the population structure of the MKRH and COOK, we applied a PCA (Figure 3), using all component breeds that were identified in the IBD analysis (Figure 2A and B). The first PC, accounting for 8.95% of variation (P = 0.048), shows separation of Arctic/sled breeds from herding breeds, with breeds historically related to the GSD intermediate to the two. The second PC accounts for 5.76% of variation but is not significant and is included here for ease of cluster visualization. The MRKH clusters closely with the HUSK and AMAL along PC1, while the COOK does not cluster with the sled dog breeds, rather situates closer to GSD and COLL along PC1. These PCA findings confirm the proportionately greater contribution of sled breeds to the MKRH, and herding breeds to the COOK, as suggested in the IBD analyses.

Figure 3.

Figure 3.

Open in a new tab

Principal component analysis plot of MKRH and COOK with putative founder breeds. PC1 accounts for 8.95% of the genetic variance of the sample set (P = 0.048). PC2 accounts for 5.76% of the variance but is not found to be significant.

We next tested for admixture in the MKRH and COOK populations using a cluster-based ancestry model which included all breeds identified from the IBD haplotype sharing analysis. To account for the close relatedness within the sample of MKRH, only the 7 least related dogs were included in this analysis, as defined above in the Methods. For breeds other than the MKRH, the remaining breed samples are assumed to be unrelated. The results are shown for K = 2–13 (Figure 4). The initial structure, arising at K = 2, separates the sled breeds from the nonsled breeds. At this value of K, the MKRH clusters with the sled breeds, whereas the COOK clusters with the nonsled breeds. The COOK separates as a distinct population at K = 5, whereas the MKRH forms a distinct cluster at K = 13. The optimal model, calculated using the CV error approach, was K = 8. At this value, the MKRH displays an average of 42.9% contribution from GREE, 24.1% from HUSK, 13.3% from AUSS, 8.3% from GSD, and <4% from the remaining breeds, indicating a clear and limited set of contributions to the modern MKRH population. The COOK population, by comparison, indicates that the dogs sampled formed their own signature at K = 5.

Figure 4.

Figure 4.

Open in a new tab

ADMIXTURE plot of admixture analysis for MKRH and COOK with putative founder breeds. Admixture analysis of K = 2–13 cluster-based ancestry models. Cross-validation analysis indicates an optimal K of 8.

Regions of Homozygosity

To identify regions of near fixation in each target breed, ROH were calculated for each dog in the COOK and MKRH populations. ROHs present in >90% of individuals from each breed underwent further analysis (Table 1). Two ROHs were identified as present in all 10 COOKs (Table 1). A region on CFA5 spanning over 326 kilobases (kb) is homozygous in all COOKs, with a section of 128 kb also homozygous for the same haplotype in AMAL, GREE, HUSK, and GSD (Supplementary Figure S1A and B). A ROH on CFA13, which spans over 1.46 Mb, is fixed for a common haplotype (Chr13_Hap_1) in all COOKs (Supplementary Figure S1C and D). This same haplotype is present in all 5 of the founder breeds, although it does not reach fixation (Figure 5).

Table 1.

ROHs with ≥95% haplotype similarity within the MKRH or COOK

Breed ROH Haplotype fixation Genes within region
COOK chr5:3919353-4295403 100% SNX19
C11orf44
COOK chr13:3003461-4464756 100% GRHL2
NCALD
HPCAL1
RRM2B
UBR5
AZIN1
KLF10
EWSR1
ODF1
RPPH1
COOK chr10:50975095-53489836 95% NRXN1
COOK chr22:60482069-61139943 95% MCF2L
F7
F10
PCID2
CUL4A
LAMP1
GRTP1
ADPRHL1
DCUN1D2
TMCO3
TFDP1
TFDP3
ATP4B
GRK1
TMEM255B
GAS6
RASA3
COOK chr29:40276838-41002224 95% GDF6
UQCRB
MTERF3
PTDSS1
SDC2
MKRH chr6:55713426-56476424 95% DR1
CCDC18
TMED5
MTF2
SNORD58B
DIPK1A
SNORA66
SNORD21
RPL5
EVI5
GFI1
Open in a new tab

Figure 5.

Figure 5.

Open in a new tab

Distribution of ROH haplotypes in MKRH and COOK and founder breeds suggests breed origins. Hierarchical clustering and breed frequency of haplotypes at ROH in COOK (A–C) and MKRH (D). Haplotypes at ROH on CFA10 and CFA22 are near fixation in COOK (A and B) and were not identified in any of the 5 proposed founder breeds. A haplotype near fixation at a COOK ROH on CFA29 (C) is absent from the AMAL, GREE, HUSK, and GSD founders, but present in the COLL, suggesting a possible inheritance of this haplotype from the COLL. The MKRH is nearing fixation for a single haplotype in a ROH on CFA6 (D). The same haplotype is also present in the AMAL, GREE, and HUSK founder breeds, but absent from the GSD and COLL, suggesting a sled dog origin of this haplotype in the MKRH. The 90% average sequence similarity between haplotypes is denoted on the dendrogram by a red dashed line. Haplotype pairs with similarity >90% are represented by the same color on the accompanying bar plot. The haplotype that is nearing fixation in the COOK or MKRH is colored light blue and indicated on the dendrogram with a blue star.

We identified 21 regions that were homozygous in >90% of the COOK chromosomes (Supplementary Table S2). Of these, 3 regions warranted additional investigation as they were either found only in COOKs, or the haplotype distribution indicated a potential source breed from amongst the founder breeds. Within each ROH, haplotypes are considered the same when their genotype similarity is >90%. A 2.51 Mb region was identified on CFA10 for which 95% of COOK chromosomes are fixed for haplotype Chr10_Hap_2/Chr10_Hap_3, which was not identified in any of the 5 founder breeds (Supplementary Figure 5A). Likewise, a 658 kb region of CFA22 demonstrated 95% fixation for haplotype Chr22_Hap_2/Chr22_Hap_3 in COOKs, but the haplotype was absent in the 5 founder breeds (Figure 5B). A region spanning 725kb on CFA29 shows a common haplotype (Chr29_Hap_1) found on 95% of COOK chromosomes. Of the 5 founder breeds, this haplotype is only found in the COLL, and at 5% frequency (Figure 5C). It is not present in any of the analyzed AMAL, GREE, HUSK dogs, or the GSD.

No ROH reached 100% in the MKRH, but one was found in 19 of 21 MKRH. Specifically, a region spanning 763 kb on CFA6 shows 95% fixation of haplotype Chr6_Hap_1/Chr6_Hap_2 in MKRH chromosomes (Figure 5D). The MKRH CFA6 haplotypes are found in 90% of GREE, 75% of AMAL, and 25% of HUSK chromosomes, but are absent from the analyzed GSDs and COLLs. The total number of haplotypes, sequence, and haplotype frequency for each ROH are given in Supplementary Table S3.

Nucleotide Diversity and Selection

Recent reports highlight the role of lipids (Hill 1974; McKenzie et al. 2008) and associated metabolism genes in dogs (Freedman et al. 2016), particularly sled dogs (Sinding 2020). We therefore analyzed nucleotide diversity (π) over 8 genes involved in regulation of fat levels and fatty acid metabolism (APOO, CERS5, FOXO1, FOXS1, MLXIPL, PDE10A, TM6SF2, and SLC25A40), all of which show signs of selection in other studies of modern sled dogs. We found reduced π over a region containing MLXIPL (Δ π = 1.23e−05 against a 95% threshold value of 1.18e−05) and TM6SF2 (Δ π = 1.07e−05 against a 95% threshold of 1.02e−05) in MKRH relative to the VILL control group. The MLXIPL gene also demonstrated decreased nucleotide diversity in the COOK compared with the VILL control (Δ π = 1.23e−05 against a 95% threshold value of 1.18e−05).

Discussion

The need for sled dogs, both to carry freight and as recreational sled dogs has, over time, led to the generation of multiple dog lines, of which two, the MKRH and COOK, are compared in this study. The history of the MKRH is complex, and both oral and written records report possible contributions from established breeds such as the Labrador retriever, NEWF, STBD, and various so called “Eskimo dogs” (http://www.sleddogcentral.com/mackenzies2.htm). These claims may reflect confusion between the MKRH, which has been bred specifically for long-distance travel, and the Mackenzie hound, which was a product of less directed breeding and used for local or personal conveyance (http://www.sleddogcentral.com/mackenzies2.htm). Our data support this distinction, showing that the MKRH ancestry includes the HUSK, GREE, and AMAL sled dog breeds, as well as the GSD and COLL, without substantial contribution from the molossoid or retriever breeds. Of note, however, we tested only one set of dogs who were themselves related samples from the premier MKRH kennel in Fairbanks, and which most likely represent the genetic makeup of a substantial portion of the modern population.

The MKRH is limited largely to Alaska and is said to be “not so much a breed as a type of dog—the freight Husky” (https://dogell.com/en/dog-breed/mackenzie-river-husky). They are reportedly able to haul large and heavy loads through deep snow and over mountainous terrains. While the lack of a breed standard portends significant variation in morphology, they require, and hence share, certain attributes including long legs, a “rangy” build, long backs, deep chests, and a slightly curving tail carried high over their back. This means they can run in a single track (foot-in-front-of-foot gait), which is optimal for freight dogs traveling long distances in deep snow. They have a strong work ethic with a pack mentality (https://dogell.com/en/dog-breed/mackenzie-river-husky) and are described as independent, intelligent, and easy to train.

Our ROH analyses revealed genes in multiple regions that support these requirements. We observe one locus on CFA6 spanning 9 genes for which 95% of the MKRH tested share the same haplotype (Figure 5D). Importantly, the MKRH CFA6 haplotypes are found in 90% of GREE, 75% of AMAL, and 25% of HUSK chromosomes, but are absent from the analyzed GSDs and COLLs. Genes of interest in the region include DIPK1A, which is associated with Aase-Smith Syndrome 2 (Soker et al. 2004; Balci, 2008), an extremely rare disorder which features red cell aplasia, and RPL5 which is associated with Diamond-Blackfan Anemia (Yu et al. 2021). Also, of interest at the 95% level in the MKRH is MTF2, a Metal Response Element Binding Transcription Factor 2 that, through a series of complex interactions, plays a key role in erythropoiesis. MTF2 is part of the PRC2.1 holocomplex of Polycomb repressive complex (PCR1), which plays key roles in cellular differentiation (Petracovici and Bonasio 2021). Of particular interest is the initial work of Rothberg et al. (2018) showing that Mtf2-PRC2 control of canonical Wnt signaling is required for erythropoiesis. Supporting this is a very recent paper showing that polycomb factor PHF19 controls cell growth and differential toward the erythroid pathway in chronic myeloid leukemia cells (Garcia-Montolio et al. 2021). The recurring theme of high levels of ROH in genes associated with red cell development in the MKRH is likely worth further investigation as increased red cell production, with its accompanying increased hemoglobin levels, offers specific advantages in terms of endurance and athleticism for freight sled dogs.

We selected the COOK as a comparison population because it is an AKC recognized freight breed but became so via a completely different route than the MKRH. While the MKRH was constructed amid limitations in breeding stock, harsh environmental selection, and functional necessity, the COOK breed was designed around the success and popularity of the dogs used in Admiral Peary’s Greenland expedition. The COOK breed was founded on Peary’s lead sled dog, a tawny-colored mastiff-type dog named Polaris, renowned for his athletic ability, gentle demeanor, and suitability as a family dog (http://www.chinookclubofamerica.org/chinook-history.html). Our data show that the COOK foundational history is comprised of the GREE and AMAL sled dog breeds, as well as the GSD and COLL herding breeds. PCA and IBD haplotype similarity shows that the primary differences between the COOK and MKRH are the inclusion of HUSK in the development of the MKRH, and a greater or more recent contribution of GSD and COLL to the COOK.

Analysis reveals 2 regions of the COOK genome that were homozygous in 100% of dogs tested (n = 10). These include a region of 326 kb on CFA5 that carries a fixed haplotype, of which 128 kb is also homozygous for the same haplotype in the AMAL, GREE, HUSK, and GSD. This region contains Sorting Nexon 19 (SNX19), which has been associated with broadly defined physiological functions such as insulin secretion, chondrogenesis, and schizophrenia (Hu et al. 2005; Kan et al. 2009; Harashima et al. 2012; Fullard et al. 2017).

We also observe a fixed region on CFA13 in COOKs which spans over 1.46 Mb that appears in all 5 founder breeds, but is not fixed, making it less applicable to this study. We also observe additional regions on CFA10, 22, and 29, which are fixed in 95% of COOK, but none were found with any significance in the founder breeds. The locus on chromosome 10, however, may be interesting and worth further investigation as it appears in many GWAS aimed at breed-specific traits (Boyko et al. 2010; Rimbault et al. 2013; Hayward et al. 2016). Genes already identified at this locus include those for body size (HMGA2) (Boyko et al. 2010; Rimbault et al. 2013) and ear structure (MSRB3) (Plassais et al. 2019). Since the locus gives such a strong signal in GWAS for a variety of structural traits, its appearance in this study makes it a strong candidate for additional investigation as to why it may be relevant for the COOK breed.

In our study, we scanned 8 lipid metabolism genes to determine whether any show changes in nucleotide diversity in either freight breed. We observed a locus spanning 600 kb and the MLXIPL gene that also showed reduced diversity against the VILL control population.

MLXIPL encodes the gene for carbohydrate response element binding protein (CHREBP) and plays several important roles. It is a glucose-responsive transcriptional factor that regulates lipid metabolism via the mTOR complex1 in diabetic nephropathy (Chen et al. 2021), and it regulates glucose tolerance and insulin sensitivity (Kursawe et al. 2013). Mutations in MLXIPL are also associated with hypertriglyceridemia in humans (Matsunaga et al. 2020). It should also be noted that MLXIPL is present within the human Williams–Beuren syndrome deletion region and occurs in a previously described homologous canine genomic region that has been shown to be under selection during domestication from wolves to dogs (Hu et al. 2005; vonHoldt et al. 2017). Williams–Beuren syndrome is characterized by hypersociability and intellectual disability, with a phenotypic variability consistent with syndromes attributed to genomic microdeletions encompassing multiple genes, as recently reviewed (Kozel et al. 2021). While there is limited evidence attributing MLXIPL to the behavioral phenotypes of Williams–Beuren syndrome (Codina-Sola et al. 2019), transposon-driven alteration of gene regulation within this region has been observed in canines (vonHoldt et al. 2017, 2018).

Our study had some limitations that should be taken into consideration when interpreting the results. Only one MKRH family was screened, and that family is obviously made up of related individuals. While we selected a subset of unrelated individuals for admixture analysis, that reduced the size of the data set significantly, yet the results appear consistent with other analysis. In addition, we screened only one comparison freight breed, the COOK. Future studies should include several other sled dog breeds. Also of note, our examination of dietary genes was restricted to those shown to be relevant for sled dogs and did not include sequence-based studies to look for variants. A more detailed list would likely offer additional insights. Finally, while many of the genes here make a provocative story, the study was done using SNPs, not whole-genome sequence and the intervals extend, in some cases, for megabases. We have no functional or statistical evidence to argue for the importance of one gene over another in any one region.

Despite these limitations, the paper revealed multiple loci for which the MKRH and the COOK, respectively, have significant ROH or decreases in nucleotide diversity, which is indicative of movement toward genomic fixation of key loci. In some cases, we can track those regions to founder breeds. Interestingly, we see little contribution to ROH on the part of the herding breeds, likely reflecting the fact that the traits under strongest selection in those breeds do not contribute to a successful sled dog. This paper serves to highlight the varied breed contributions of 2 functionally bred dog populations with overlapping physical requirements and lays the foundation for identification of genomic regions of both shared and divergent relevance to sled dogs.

Supplementary Material

esac002_suppl_Supplementary_Figure_S1
Click here for additional data file. (1.9MB, pdf)
esac002_suppl_Supplementary_Table_S1
Click here for additional data file. (24.8KB, docx)
esac002_suppl_Supplementary_Table_S2
Click here for additional data file. (15.3KB, docx)
esac002_suppl_Supplementary_Table_S3
Click here for additional data file. (22.1KB, docx)

Acknowledgments

We thank Andrew Hogan and Drs. Heidi G. Parker and Jacquelyn Evans for helpful suggestions and mentoring. We also thank Dr. Heather Huson for sharing expertise regarding sled dogs. We thank the NIH and NHGRI Offices of Intramural Training and Education for assistance in writing this manuscript. Finally, we thank dog owners, particularly Donna Dowling, for sharing DNA samples from their dogs. The authors state that they have nothing to declare and no conflict of interest. The sponsors of this work had no role in the study design or analysis. The paper has not been published in whole or part elsewhere.

Contributor Information

Muhammad Basil Ali, National Institute for Biotechnology and Genetic Engineering College (NIBGE), Faisalabad, Pakistan; Pakistan Institute of Engineering and Applied Sciences (PIEAS), Islamabad, Pakistan.

Dayna L Dreger, National Human Genome Research Institute, National Institutes of Health, 50 South Drive, Building 51, Room 5351, Bethesda, MD 20892, USA.

Reuben M Buckley, National Human Genome Research Institute, National Institutes of Health, 50 South Drive, Building 51, Room 5351, Bethesda, MD 20892, USA.

Shahid Mansoor, National Institute for Biotechnology and Genetic Engineering College (NIBGE), Faisalabad, Pakistan; Pakistan Institute of Engineering and Applied Sciences (PIEAS), Islamabad, Pakistan.

Qaiser M Khan, National Institute for Biotechnology and Genetic Engineering College (NIBGE), Faisalabad, Pakistan; Pakistan Institute of Engineering and Applied Sciences (PIEAS), Islamabad, Pakistan.

Elaine A Ostrander, National Human Genome Research Institute, National Institutes of Health, 50 South Drive, Building 51, Room 5351, Bethesda, MD 20892, USA.

Funding

This work was supported by the Intramural Program of the National Human Genome Research Institute of the National Institutes of Health (NIH) (E.A.O., D.L.D., R.M.B.). While a graduate student at NIH, M.B.A. was similarly funded by NIH.

Author Contributions

Conceptualization: E.A.O. Data curation: M.B.A., R.M.B. Formal analysis: M.B.A., D.L.D. Funding acquisition: E.A.O., S.M. Investigation: M.B.A., D.L.D. Methodology: M.B.A., D.L.D., R.M.B. Project administration: D.L.D., E.A.O. Resources: E.A.O. Supervision: E.A.O., Q.M.K., S.M. Validation: D.L.D.

Visualization: M.B.A., R.M.B. Writing—original draft: M.B.A., D.L.D., E.A.O. Writing—review and editing: M.B.A., D.L.D., E.A.O., S.M., Q.M.K.

Data Availability

Sampling locations and primer sequences are contained within the paper and supplement. Relationship of MKRH to one another is shown in Figure 1. Description of COOK is contained within the methods. Precise nucleotide locations are in Table 1 and Supplementary Table S2.

References

  1. Abraham  G, Qiu Y, Inouye M.  2017. FlashPCA2: principal component analysis of Biobank-scale genotype datasets. Bioinformatics. 33:2776–2778. [DOI] [PubMed] [Google Scholar]
  2. Alexander  DH, Lange K.  2011. Enhancements to the ADMIXTURE algorithm for individual ancestry estimation. BMC Bioinf. 12:246. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Ameen  C, Feuerborn TR, Brown SK, Linderholm A, Hulme-Beaman A, Lebrasseur O, Sinding MS, Lounsberry ZT, Lin AT, Appelt M, et al.  2019. Specialized sledge dogs accompanied Inuit dispersal across the North American Arctic. Proc Biol Sci. 286:20191929. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. American Kennel Club. 2017. The new complete dog book. Mount Joy (PA): Fox Chapel Publishing. [Google Scholar]
  5. Balci  S.  2008. Limb anomalies and anemia: Aase-Smith syndrome. Fetal Pediatr Pathol. 27:292–293. [DOI] [PubMed] [Google Scholar]
  6. Boyko  AR, Quignon P, Li L, Schoenebeck JJ, Degenhardt JD, Lohmueller KE, Zhao K, Brisbin A, Parker HG, vonHoldt BM, et al.  2010. A simple genetic architecture underlies morphological variation in dogs. PLoS Biol. 8:e1000451. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Browning  SR, Browning BL.  2015. Accurate non-parametric estimation of recent effective population size from segments of identity by descent. Am J Hum Genet. 97:404–418. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chen  N, Mu L, Yang Z, Du C, Wu M, Song S, Yuan C, Shi Y.  2021. Carbohydrate response element-binding protein regulates lipid metabolism via mTOR complex1 in diabetic nephropathy. J Cell Physiol. 236:625–640. [DOI] [PubMed] [Google Scholar]
  9. Codina-Sola  M, Costa-Roger M, Pérez-García D, Flores R, Palacios-Verdú MG, Cusco I, Pérez-Jurado LA.  2019. Genetic factors contributing to autism spectrum disorder in Williams-Beuren syndrome. J Med Genet. 56:801–808. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Ford  J.A.  1959. Eskimo prehistory in the vicinity of Point Barrow, Alaska. Anthropol Pap Am Mus Nat Hist. 47:1. [Google Scholar]
  11. Freedman  AH, Schweizer RM, Ortega-Del Vecchyo D, Han E, Davis BW, Gronau I, Silva PM, Galaverni M, Fan Z, Marx P, et al.  2016. Demographically-based evaluation of genomic regions under selection in domestic dogs. PLoS Genet. 12:e1005851. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fullard  JF, Giambartolomei C, Hauberg ME, Xu K, Voloudakis G, Shao Z, Bare C, Dudley JT, Mattheisen M, Robakis NK, Haroutunian V, Roussos P.  2017. Open chromatin profiling of human postmortem brain infers functional roles for non-coding schizophrenia loci. Hum Mol Genet. 26:1942–1951. doi: 10.1093/hmg/ddx103 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. García-Montolio  M, Ballaré C, Blanco E, Gutiérrez A, Aranda S, Gómez A, Kok CH, Yeung DT, Hughes TP, Vizán P, et al.  2021. Polycomb factor PHF19 controls cell growth and differentiation toward erythroid pathway in chronic myeloid leukemia cells. Front Cell Dev Biol. 9:655201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Harashima  S, Horiuchi T, Wang Y, Notkins AL, Seino Y, Inagaki N.  2012. Sorting nexin 19 regulates the number of dense core vesicles in pancreatic β-cells. J Diabetes Investig. 3:52–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hayward  JJ, Castelhano MG, Oliveira KC, Corey E, Balkman C, Baxter TL, Casal ML, Center SA, Fang M, Garrison SJ, et al.  2016. Complex disease and phenotype mapping in the domestic dog. Nat Commun. 7:10460. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hill  WG.  1974. Estimation of linkage disequilibrium in randomly mating populations. Heredity (Edinb). 33:229–239. [DOI] [PubMed] [Google Scholar]
  17. Hu  YF, Zhang HL, Cai T, Harashima S, Notkins AL.  2005. The IA-2 interactome. Diabetologia. 48:2576–2581. [DOI] [PubMed] [Google Scholar]
  18. Kan  A, Ikeda T, Saito T, Yano F, Fukai A, Hojo H, Ogasawara T, Ogata N, Nakamura K, Chung UI, et al.  2009. Screening of chondrogenic factors with a real-time fluorescence-monitoring cell line ATDC5-C2ER: identification of sorting nexin 19 as a novel factor. Arthritis Rheum. 60:3314–3323. [DOI] [PubMed] [Google Scholar]
  19. Kopelman  NM, Mayzel J, Jakobsson M, Rosenberg NA, Mayrose I.  2015. Clumpak: a program for identifying clustering modes and packaging population structure inferences across K. Mol Ecol Resour. 15:1179–1191. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kozel  BA, Barak B, Kim CA, Mervis CB, Osborne LR, Porter M, Pober BR.  2021. Williams syndrome. Nat Rev Dis Primers. 7:42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kursawe  R, Caprio S, Giannini C, Narayan D, Lin A, D’Adamo E, Shaw M, Pierpont B, Cushman SW, Shulman GI.  2013. Decreased transcription of ChREBP-α/β isoforms in abdominal subcutaneous adipose tissue of obese adolescents with prediabetes or early type 2 diabetes: associations with insulin resistance and hyperglycemia. Diabetes. 62:837–844. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Librado  P, Rozas J, 2009. DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics. 25:1451–1452. [DOI] [PubMed] [Google Scholar]
  23. Lindblad-Toh  K, Wade CM, Mikkelsen TS, Karlsson EK, Jaffe DB, Kamal M, Clamp M, Chang JL, Kulbokas EJ 3rd, Zody MC, et al.  2005. Genome sequence, comparative analysis and haplotype structure of the domestic dog. Nature. 438:803–819. [DOI] [PubMed] [Google Scholar]
  24. Matsunaga  A, Nagashima M, Yamagishi H, Saku K.  2020. Variants of lipid-related genes in adult Japanese patients with severe hypertriglyceridemia. J Atheroscler Thromb. 27:1264–1277. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. McKenzie  EC, Hinchcliff KW, Valberg SJ, Williamson KK, Payton ME, Davis MS.  2008. Assessment of alterations in triglyceride and glycogen concentrations in muscle tissue of Alaskan sled dogs during repetitive prolonged exercise. Am J Vet Res. 69:1097–1103. [DOI] [PubMed] [Google Scholar]
  26. Morey  DF, Aaris-Sorensen K, 2002. Paleoeskimo dogs of the Eastern Arctic. Arctic. 55, 44–56. [Google Scholar]
  27. Parker  HG, Dreger DL, Rimbault M, Davis BW, Mullen AB, Carpintero-Ramirez G, Ostrander EA.  2017. Genomic analyses reveal the influence of geographic origin, migration, and hybridization on modern dog breed development. Cell Rep. 19:697–708. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Petracovici  A, Bonasio R.  2021. Distinct PRC2 subunits regulate maintenance and establishment of Polycomb repression during differentiation. Mol Cell. 81:2625–2639.e5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Pitulko  VV, Kasparav AK.  2017. Archaeological dogs from the early Holocene Zhokhov site in the Eastern Siberian Arctic. J Arch Sci. 13:491–515. [Google Scholar]
  30. Plassais  J, Kim J, Davis BW, Karyadi DM, Hogan AN, Harris AC, Decker B, Parker HG, Ostrander EA.  2019. Whole genome sequencing of canids reveals genomic regions under selection and variants influencing morphology. Nat Commun. 10:1489. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Purcell  S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D, Maller J, Sklar P, de Bakker PI, Daly MJ, et al.  2007. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 81:559–575. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Quinlan  AR, Hall IM.  2010. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics. 26:841–842. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Rimbault  M, Beale HC, Schoenebeck JJ, Hoopes BC, Allen JJ, Kilroy-Glynn P, Wayne RK, Sutter NB, Ostrander EA.  2013. Derived variants at six genes explain nearly half of size reduction in dog breeds. Genome Res. 23:1985–1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Rosenthal  EA, Ranchalis J, Crosslin DR, Burt A, Brunzell JD, Motulsky AG, Nickerson DA, Wijsman EM, Jarvik GP; NHLBI GO Exome Sequencing Project. 2013. Joint linkage and association analysis with exome sequence data implicates SLC25A40 in hypertriglyceridemia. Am J Hum Genet. 93:1035–1045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Rothberg  JLM, Maganti HB, Jrade H, Porter CJ, Palidwor GA, Cafariello C, Battaion HL, Khan ST, Perkins TJ, Paulson RF, et al.  2018. Mtf2-PRC2 control of canonical Wnt signaling is required for definitive erythropoiesis. Cell Discov. 4:21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Sambrook  J, Fritsch EF, Maniatis T.  1989. Isolation of high-molecular weight DNA from mammalian cells. In: Nolan C, editor. Molecular cloning: a laboratory manual. Cold Spring Harbor (NY): Cold Spring Harbor Press. p. 9.16–19.19. [Google Scholar]
  37. Serpell  JA, 1995. The domestic dog: its evolution, behavior and interactions with people, 1st ed. Cambridge (UK): Cambridge University Press. [Google Scholar]
  38. Shannon  LM, Boyko RH, Castelhano M, Corey E, Hayward JJ, McLean C, White ME, Abi Said M, Anita BA, Bondjengo NI, et al.  2015. Genetic structure in village dogs reveals a Central Asian domestication origin. Proc Natl Acad Sci USA. 112:13639–13644. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Sinding  MS, Gopalakrishnan S, Ramos-Madrigal J, de Manuel M, Pitulko VV, Kuderna L, Feuerborn TR, Frantz LAF, Vieira FG, Niemann J, et al.  2020. Arctic-adapted dogs emerged at the Pleistocene-Holocene transition. Science. 368:1495–1499. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Soker  M, Ayyildiz O, Isikdogan A.  2004. Aase-Smith syndrome type II. Saudi Med J. 25:2004–2006. [PubMed] [Google Scholar]
  41. Talenti  A, Dreger DL, Frattini S, Polli M, Marelli S, Harris AC, Liotta L, Cocco R, Hogan AN, Bigi D, Caniglia R, Parker HG, Pagnacco G, Ostrander EA, Crepaldi P.  2019. Studies of modern Italian dog populations reveal multiple patterns for domestic breed evolution. Ecol Evol. 8:2911–2925. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Turkieh  A, Caubère C, Barutaut M, Desmoulin F, Harmancey R, Galinier M, Berry M, Dambrin C, Polidori C, Casteilla L, et al.  2014. Apolipoprotein O is mitochondrial and promotes lipotoxicity in heart. J Clin Invest. 124:2277–2286. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. vonHoldt  BM, Ji SS, Aardema ML, Stahler DR, Udell MAR, Sinsheimer JS.  2018. Activity of genes with functions in human Williams-Beuren syndrome is impacted by mobile element insertions in the gray wolf genome. Genome Biol Evol. 10:1546–1553. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. vonHoldt  BM, Shuldiner E, Koch IJ, Kartzinel RY, Hogan A, Brubaker L, Wanser S, Stahler D, Wynne CDL, Ostrander EA, et al.  2017. Structural variants in genes associated with human Williams-Beuren syndrome underlie stereotypical hypersociability in domestic dogs. Sci Adv. 3:e1700398. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Wendt  R.  1999. Alaska dog mushing guide: facts and legends. 5th ed.Fairbanks (AK): Goldstream Publications,. [Google Scholar]
  46. Yu  L, Lemay P, Ludlow A, Guyot MC, Jones M, Mohamed FF, Saroya GA, Panaretos C, Schneider E, Wang Y, et al.  2021. A new murine Rpl5 (uL18) mutation provides a unique model of variably penetrant Diamond-Blackfan anemia. Blood Adv. 5:4167–4178. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Zhao  S, Jing W, Samuels DC, Sheng Q, Shyr Y, Guo Y.  2018. Strategies for processing and quality control of Illumina genotyping arrays. Brief Bioinform. 19:765–775. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

esac002_suppl_Supplementary_Figure_S1
Click here for additional data file. (1.9MB, pdf)
esac002_suppl_Supplementary_Table_S1
Click here for additional data file. (24.8KB, docx)
esac002_suppl_Supplementary_Table_S2
Click here for additional data file. (15.3KB, docx)
esac002_suppl_Supplementary_Table_S3
Click here for additional data file. (22.1KB, docx)

Data Availability Statement

Sampling locations and primer sequences are contained within the paper and supplement. Relationship of MKRH to one another is shown in Figure 1. Description of COOK is contained within the methods. Precise nucleotide locations are in Table 1 and Supplementary Table S2.

Articles from Journal of Heredity are provided here courtesy of Oxford University Press

RESOURCES