Whole Genome Association Study to Detect Single Nucleotide Polymorphisms for Behavior in Sapsaree Dog (Canis familiaris)

Article information

Asian-Australas J Anim Sci. 2015;28(7):936-942
1School of Life Science, Kyungpook National University, Daegu, Korea.
*Corresponding Author: J.-J. Kim. Tel: +82-53-810-3027, Fax: +82-53-801-3027, E-mail: kimjj@ynu.ac.kr
School of Biotechnology, Yeungnam University, Gyeongsan 712-749, Korea
aThe two authors equally contributed to this work
Received 2014 December 15; Revised 2015 January 06; Accepted 2015 April 01.

Abstract

The purpose of this study was to characterize genetic architecture of behavior patterns in Sapsaree dogs. The breed population (n = 8,256) has been constructed since 1990 over 12 generations and managed at the Sapsaree Breeding Research Institute, Gyeongsan, Korea. Seven behavioral traits were investigated for 882 individuals. The traits were classified as a quantitative or a categorical group, and heritabilities (h2) and variance components were estimated under the Animal model using ASREML 2.0 software program. In general, the h2 estimates of the traits ranged between 0.00 and 0.16. Strong genetic (rG) and phenotypic (rP) correlations were observed between nerve stability, affability and adaptability, i.e. 0.9 to 0.94 and 0.46 to 0.68, respectively. To detect significant single nucleotide polymorphism (SNP) for the behavioral traits, a total of 134 and 60 samples were genotyped using the Illumina 22K CanineSNP20 and 170K CanineHD bead chips, respectively. Two datasets comprising 60 (Sap60) and 183 (Sap183) samples were analyzed, respectively, of which the latter was based on the SNPs that were embedded on both the 22K and 170K chips. To perform genome-wide association analysis, each SNP was considered with the residuals of each phenotype that were adjusted for sex and year of birth as fixed effects. A least squares based single marker regression analysis was followed by a stepwise regression procedure for the significant SNPs (p<0.01), to determine a best set of SNPs for each trait. A total of 41 SNPs were detected with the Sap183 samples for the behavior traits. The significant SNPs need to be verified using other samples, so as to be utilized to improve behavior traits via marker-assisted selection in the Sapsaree population.

INTRODUCTION

Among the domestic animals, the dog takes a significant position in human society due to its very intimate behavioral exhibition with humans. Moreover, the dog is one of the most diverse domestic species in terms of morphology and behavior (Wayne and Ostrander, 1999). Although each breed shows general uniformity for behavior and morphology, individuals within a breed have diverse heritages because of different haplotypes related to the traits (Vila et al., 1997), which can be supported by analysis of protein alleles (Ferrell et al., 1978), as well as with hypervariable microsatellite loci (Fredholm and Wintero, 1995).

Currently, there are more than 400 registered dog breeds around the world that have been bred for various purposes, e.g. hunting, guarding, guides for blind, pets, etc. The Sapsaree is one of the aboriginal breeds in Korea, with medium body size and body height ranging 49 to 55 cm (Kim et al., 2001). Adult coat hair is long and abundant with two typical variations in color, i.e. blue and yellow. The Sapsaree dogs are very gentle, protective and loyal to their owner. Generally, the dogs are not aggressive, but express aggression, if other dogs enter their territory. The Sapsaree population was close to extinction during Japanese colonization. Afterwards, in 1986, eight individuals with similar color and body conformation to Sapsaree breed were collected across the country by local Sapsaree lovers in Daegu, Korea. Successively, systematic mating and reproduction generated the current population of about 3,000 individuals including five hundred at the Sapsaree Breeding Research Institute in Gyeongsan, Gyeongbuk province.

Recent sequencing technologies allowed the sequencing of the entire canine genome (Eggen, 2012), which provides abundant genetic resources to search for genetic variants underlying diseases and phenotypes such as body size (Patterson et al., 1982; Ostrander et al., 1997; Galibert et al., 1998). Whole genome association (WGA) studies are now routinely practiced due to the availability of high density canine single nucleotide polymorphism (SNP) chips such as the Illumina SNP arrays (Karlsson and Lindblad-Toh, 2008; Spady and Ostrandar, 2008). Further, most dog breeds are less than 200 years old, and, thus, have high linkage disequilibrium (LD) and long haplotype blocks (Sutter et al., 2004), which enables WGA study with lower density marker maps, compared to humans (Sutter and Ostrander, 2004; Karlsson and Lindblad-Toh, 2008). According to Karlsson and Lindblad-Toh (2008), about 15,000 SNPs proved to be sufficient for WGA mapping. After the first report of a WGA study in a dog by Karlsson et al. (2007), numerous WGA studies in dogs have been carried out for disease traits. However, there were few reports about GWA analysis on morphology and behavior.

The objectives of this study were to genetically characterize Sapsaree breed and to detect SNPs or quantitative trait loci (QTL) for various behavior traits by WGA analysis in the Sapsaree population.

MATERIALS AND METHODS

Pedigree and phenotypes

The pedigree of the Sapsaree population (n = 8,256) was constructed from 1989 to 2007. A set of 1,014 individuals were recorded for morphology and behavior traits. The behavior tests were first introduced in 1998 and carried out once a year. Each animal was tested only once for nerve stability (NST), affability (AFB), wariness (WRN), adaptability (ADP), sharpness (SRP), activity (ACT) or energy level or temperament and reactions during blood drawn (RBD). Behavior definitions and measurements are described in Tables 1 and 2. All the individuals passed through a sufficient adjustment period before they underwent behavior tests and, the behavior measures were recorded mainly by one experimenter. Dog’s behaviors were graded according to intensities, for which the lowest to the greatest scores reflected the least to the most desired expression, respectively.

The definition of behaviors traits used in the experiment

Description of traits as quantitative and categorical measurements

Variance component estimation

The behavior traits that were recorded at hedonic scales were analyzed with either quantitative or continuous measurements. For Some behavior traits, the records were transformed to binary pattern, i.e. absence (0) or presence (1). To do that, two or more categories were abridged together (Table 2). Sex, season of birth (summer and winter), birth year groups (10 levels) and age at testing (<15, 20, 25, and, >25 months) were considered as fixed factors for analysis. Then, the SAS general linear model (GLM) procedure (Release 9.1, SAS Institute Inc., NC, USA, 1999) was applied to determine the fixed effects to be fitted in the Animal model at α = 0.1 level.

A restricted maximum likelihood approach was applied under the Animal model using ASReml 2.0 software (Gilmour et al., 2001). A univariate analysis for each trait with quantitative measurements was conducted to estimate variance components, direct and maternal heritability (Equation 1)

(Equation 1) Y=μ+(si+sbj+ybk+tal)+(am+mn)+eijklmn

Where, Y = phenotype, μ = overall mean, si = effect of sex, sbj = effect of season of birth, ybk = effect of year of birth; tal = effect of age at testing, am = animal effect with N(0, Aσ2d), where A = the additive relationship matrix; mn = maternal genetic effect with N(0, Aσ2m); eijklmn = the random error term with N(0, Iσ2e), where I = the identity matrix. σ2d, σ2m, and σ2e are direct and maternal genetic and residual (error term) variances, respectively. No interaction (covariance component) between maternal and direct effects was assumed. Direct and maternal heritabilities were defined as hd2=σd2σp2 and hm2=σm2σp2, where σp2 is phenotypic variance such that σp2=σd2+σm2+σe2. An alternative model was applied without the maternal effect (Equation 2),

(Equation 2) Y=μ+si+sbj+ybk+tal+am+eijklm

in which the overall heritability was: hoverall2=σa2σp2 where, σp2=σa2+σe2. σa2 is additive genetic variance. Under the model, phenotypic and genetic correlations were calculated using ASReml 2.0, according to Falconer and Mackay (1996). Multivariate analyses were performed to estimate genetic and environmental correlations between the behavior traits.

The behavior traits with binary data formats were also modeled with a generalized linear mixed model (GLMM) using ‘!logit’ function and then fitted under an Animal model (Equation 1). The heritabilities for categorical traits with binary scale (0 or 1) were estimated by the Equation 3 (Lynch and Walsh, 1998):

(Equation 3) h(0.1)2=σA(0.1)2Φ_p(1-Φ_p)

Where, σA(0,1)2=additive genetic variance, Φp = incidence in the population and, Φp (1−Φp) = phenotypic variance on the observed scale. However, a multivariate analysis with this binary data set was not available due to the lack of fit for appropriate logistic models, when ASReml 2.0 was applied. Thus, correlation estimates between the traits only with quantitative measures were obtained.

Genome-wide association test

Genotyping was performed using two types of canine SNP chips, i.e. CanineSNP20 BeadChip and CanineHD BeadChip. The former and the latter contained more than 22,000 and 170,000 evenly spaced and validated SNPs derived from the CanFam2.0 assembly. Initially, a total of 134 individuals were genotyped by CanineSNP20 BeadChip using the Illumina’s Infinium Assay. Another 60 individuals were genotyped with CanineHD BeadChip using the Infinium HD Assay Protocol (Illumina Inc., San Diego, CA, USA). The genotyping analyses were performed at GeneSeek (Ltd.) and the SNP data were obtained.

A thorough screening of the SNP were performed on three different data; Sap134, Sap60, and Sap183, of which the first two data sets were based on CanineSNP20 and CanineHD BeadChip panels, respectively, and the last set was derived from the SNPs that were embedded on both SNP chips. The genotyped animals were excluded if any particular genotype(s) was entirely missed. No SNP with <0.05 minor allele frequency or with more than 10% missing genotype was included in the following association tests. The genome-wide association tests were carried out mainly on 38 autosomes with the Sap183 data set, because of its greater sample size compared to the Sap60 and Sap134.

For association studies with quantitative traits, sex and year of birth was fitted as fixed factors, and the residuals for each trait were obtained in the process. A generalized linear model procedure was implemented using PROC GLM in SAS statistical software (SAS version 9.1, SAS Inst., Inc., Cary, NC, USA) to calculate the error variances.

In a random mating population with no population structure the association between a marker and a trait can be tested with single marker regression as:

Y=1n+βaXa+βdXd+e

Where, Y is a vector of phenotypes residuals for any given trait, 1n is a vector of 1s, Xa and Xd are coefficient matrices allocating records to the additive and dominance effects of the SNP, respectively, βa and βd are the respective coefficients of the marker effect, and e is a vector of random deviates, N(0, σe2), where σ2e is the error variance. For Xa (Xd), 1, 0, and −1 (0, 1, 0) were assigned for the SNP genotypes, AA, AB, and BB, respectively. The null hypothesis is that the marker has no effect on the trait, while the alternative hypothesis is that the marker does affect the trait, due to LD of the SNP with a QTL.

The variation explained by each SNP (S2SNP) was calculated as i=13αi2fi-μ2, where i denotes each genotype, αi is allele substitution effect (= −a, d, and +a for BB, AB, and AA, respectively, in which a and d were estimated from the simple regression analysis for the SNP), fi is the frequency of ith genotype, μ is the population mean that can be expressed as (fAA-fBB)a+fABd (Falconer and Mackay, 1996). Proportion of phenotypic variance due to the SNP was then estimated as S2SNP/S2P, in which S2P was obtained from residual values of the trait after adjusting fixed effects. Therefore, the estimate of the proportion of phenotype variance due to all of the significant SNPs was ∑S2SNPi/S2P.

To determine significant SNP for the behavior traits, a significance threshold of 1% point-wise p value from F distribution was applied for each trait. Then, a best set of significant SNPs were selected from the SNPs using a stepwise regression procedure (Neter et al., 1990), due to the fact that some of the significant SNPs, if closely linked to each other (LD), would yield redundant information in implementing a marker assisted selection program. Inclusion and exclusion of each SNP out of the stepwise model was determined at 0.001 level.

RESULTS AND DISCUSSION

Heritability and correlations between behavior traits

Heritability estimates (h2) for behavior traits were presented in Table 3 and 4. The h2 estimates fell within a low range (0.00 to 0.17). In general, there was no noticeable difference of the estimates between the single and multiple trait analyses, except for AFB and ACT. All behavior traits except RBD did not have a great maternal h2 estimate. Nerve stability, ADP and RBD had low direct h2 estimates, i.e. close to zero.

Heritability estimates of behavior traits with quantitative values

Heritability estimates of behavior traits with binary values

Famula (2001) reported that heritability of dog behaviors were low to high (0.10 to 0.60) depending on the characters and breeds. Wilsson and Sundgren (1997), Van der Waaij et al. (2008), and Ruefenacht et al. (2002) reported a low to medium h2 for NST (0.15 to 0.25), AFB (0.03 to 0.38), and SRP (0.09 to 0.19) of dogs. For ACT trait studied by Wilsson and Sundgren (1997) and related behavior traits such as temperament by van der Waaij et al. (2008) and Ruefenacht et al. (2002) and energy by Bartlett (1976), the h2s ranged from 0.05 to 0.53. Reuterwall and Ryman (1973) reported h2s of 0.09 to 0.17 for AFB and 0.00 to 0.04 for ADP, respectively. Goddard and Beilharz (1982) estimated h2 of 0.10 for suspicion in Labrador retrievers. Our estimates were, in general, concordant with the above reports. Ruefenacht et al. (2002) reported that h2 estimates under quantitative measures for behaviors were lower or similar to those of categorical estimates, which was consistent with this study (Tables 3 and 4).

Wilsson and Sundgren (1997) reported small or negligible maternal effects for behaviors in old dogs. Our results showed that, in general, the maternal effects for behavior with quantitative measures were small, except for RBD (Table 3), but were moderate for AFB, ADP, and RBD with binary values (Table 4).

Genetic correlations (rG) were, in general, greater than phenotypic correlations (rP) (Table 5). Moderate positive rPs were observed between NST and AFB (0.68), between NST and ADP (0.6) and between AFB and ADP (0.62). There were strong genetic correlations between all the behavior traits, except for RBD (Table 5). More specifically, NST, AFB, WRN, and ADP had strong genetic correlation, while RBD had negative genetic correlations with other behavior traits, e.g. with ACT (−0.52).

Phenotypic (lower diagonal) and genetic correlation (upper diagonal) of behavior traits with quantitative values

Ruefenacht et al. (2002) reported positive rPs (0.31 to 0.57) and rGs (0.34 to 0.83) among NST, SRP, and temperament (or ACT in this study) in German Shepherd dogs. Van der Waaij et al. (2008) also reported a range from 0.12 to 0.36 (rP), and −0.51 to 0.64 (rG) among the comparable traits (i.e., NST, AFB, SRP, and ACT) in this study, with the greatest rP (0.36) between NST and temperament, and the greatest rG (0.64) between NST and AFB in GS dogs, which was similar to the results in this study (Table 5).

Significant single nucleotide polymorphisms for behavior by whole genome association studies

For the Sap183 data, a set of 15,825 SNPs were chosen after the quality control tests. The number of available SNPs on each chromosome was proportionate to its chromosomal length, e.g. 865 SNPs on CFA1 and 199 SNPs on CFA38, respectively (results not shown). The physical map of SNPs spanned about 2,181 Mbps, with an average distance of 137.8±147.4 Kb between adjacent SNPs.

A total of 41 significant SNPs for the seven behavior traits were determined from the stepwise regression analyses (Table 6). The set of SNPs for each trait explained a large proportion of total phenotypic variance (29% to 67%), partly due to the fact that the SNP effect was overestimated with the small sample size in this study (n = 183). In general, the SNP effects for the behavior measures had both additive and dominance effects. However, WRN and SRP had mainly additive and dominance effects, respectively (Table 6). The SNPs were distributed across the canine genome, i.e. 24 chromosomes (CFA). The greatest number (6) of SNPs was detected on CFA18, among which four SNPs for NST, AFB, ADP, and ACT were located at 23 or 26 Mbs (Table 6). The significant SNPs in this study need to be verified using other samples, so as to be utilized to improve behavior traits via marker-assisted selection in the Sapsaree population.

Identities, positions, and effects of the SNPs associated with behavior traits in a Sapsaree population using Sap183 SNP panel

ACKNOWLEDGMENTS

This research was supported by the KNU research fund in 2012 and by the Technology Development Program for Agriculture and Forestry, Ministry of Agriculture, Forestry, and Fisheries, Republic of Korea, 2011.

References

Bartlett CR. 1976. Heritabilities and Genetic Correlations between Hip Dysplasia and Temperament Traits of Seeing-eyedogs. Masters thesis Rutgers University. NewBrunswick, NJ, USA:
Eggen A. 2012;The development and application of genomic selection as a new breeding paradigm. Anim Front 2:10–15.
Falconer DS, Mackay TFC. 1996. Introduction to Quantitative Genetics 4th ednth ed. Longman Group Ltd.. Essex, England:
Famular TR. 2001. Genetics of quantitative traits and improvement of dog breeds. The Genetics of the Dog In : Ruvinsky A, Sampson J, eds. CABI Publishing, CAB International. Wallingford, Oxon, UK: p. 487–503.
Ferrell RE, Morizot DC, Horn J, Carley CJ. 1980;Biochemical markers in species endangered by introgression: The red wolf. Biochem Genet 18:39–49.
Fredholm M, Wintero AK. 1995;Variation of short tandem repeats within and between species belonging to the Canidae family. Mamm Genome 6:11–18.
Galibert F, Andre C, Cheron A, Chuat JC, Hitte C, Jiang Z, Jouquand S, Priat C, Renier C, Vignaux F. 1998;The importance of the canine model in medical genetics. Bull Acad Natl Med 182:811–821.
Gilmour AR, Cullis BR, Welham SJ, Thompson R. 2001. ASREML Reference Manual NSW Agriculture. Orange, Australia:
Goddard ME, Beilharz RG. 1982;Genetics of traits which determine the suitability of dogs as guide-dogs for the blind. Appl Anim Ethol 9:299–315.
Han KI, Alam M, Lee YM, Lee DH, Ha JH, Kim JJ. 2010;A study on t he morphology and Behavior of the Sapsaree: A Korean native dog (Canis familiaris). J Anim Sci Technol 52:481–490.
Karlsson EK, Baranowska I, Wade CM, Salmon Hillbertz NH, Zody MC, Anderson N, Biagi TM, Patterson N, Pielberg GR, Kulbokas EJ 3rd, et al. 2007;Efficient mapping of mendelian traits in dogs through genome-wide association. Nat Genet 39:1321–1328.
Karlsson EK, Lindblad-Toh K. 2008;Leader of the pack: gene mapping in dogs and other model organisms. Nat Rev Genet 9:713–725.
Kim KS, Jeong HW, Park CK, Ha JH. 2001;Suitability of AFLP markers for the study of Genetic relationships among Korean native dogs. Genes Genet Syst 76:243–250.
Lynch M, Walsh B. 1998. Genetics and analysis of quantitative traits Sinauer. Sunderland, MA, USA:
Neter J, Wasserman W, Kutner MH. 1990. Applied Linear Statistical Models 3rd edth ed. McGraw-Hill/Irwin. Boston, MA, USA:
Ostrander EA, Giniger E. 1997;Semper fidelis: What man’s best friend can teach us about human biology and disease. Am J Hum Genet 61:475–480.
Patterson DF, Haskins ME, Jezyk PF. 1982;Models of human genetic disease in domestic animals. Adv Hum Genet 12:263–339.
Reuterwall C, Ryman N. 1973;An estimate of the magnitude of additive genetic variation of some mental characters in Alsatian dogs. Hereditas 73:277–283.
Ruefenacht S, Gebhardt-henrich S, Miyake T, Gaillard C. 2002;A behavior test of German Shepherd dogs: heritability of seven different traits. Appl Anim Behav Sci 79:113–132.
Spaady TC, Ostrander EA. 2008;Canine behavioral genetics: Pointing out the phenotypes and herding up the genes. Am J Hum Genet 82:10–18.
Sutter NB, Bustamante CD, Chase K, Gray MM, Zhao K, Zhu L, Padhukasahasram B, Karlins E, Davis S, Jones PG, Quignon P, Johnson GS, Parker HG, Fretwell N, Mosher DS, Lawler DF, Satyaraj E, Nordborg M, Lark KG, Wayne RK, Ostrander EA. 2007;A single IGFI allele is a major determinant of small size in dogs. Science 316(5821):112–115.
Sutter NB, Ostrander EA. 2004;Dog star rising: The canine genetic system. Nat Rev Genet 5:900–910.
Vila C, Savolainen P, Maldonado JE, Amorim IR, Rice JE, Honeycut RL, Crandall KA, Lundeberg J, Wayne RK. 1997;Multiple and ancient origins of the domestic dog. Science 276(5319):1687–1689.
Wayne RK, Ostrander EA. 1999;Origin, genetic diversity, and genome structure of the domestic dog. BioEssays 21:247–257.
Wilsson E, Sundgren P-E. 1997;The use of a behavior test for selection of dogs for service and breeding. II: Heritability for tested parameters and effect of selection based on service dog characteristics. Appl Anim Behav Sci 54:235–241.

Article information Continued

Table 1

The definition of behaviors traits used in the experiment

Trait* Definition
Nerve stability (NST) 1 The appropriateness of the dog’s reaction to a certain situation. This includes the dog’s ability to adapt to various types of situations, to concentrate when highly aroused or in a situation of conflict, as well as its ability to relax and to overcome a frightening situation.
Affability (AFB) 1 The dog’s willingness to make contact with people.
Wariness (WRN) The dogs’ approach to be cautious about the probable upcoming threat.
Adaptability (ADP) Coping ability of the dog’s with changes in the physical or natural environment.
Activity/energy level/temperament (ACT) 1 The degree of liveliness. Dogs with high activity or temperament are more responsive to all types of stimuli. 2 In other words, temperament is the physical flexibility and intensity of reactions to environment with a big radius of action.
Sharpness (SRP) 2 The ability to react in an aggressive way towards a serious or serious looking attack. The sharpness is desired only when the dog is threatened. After the end of the threatening period, the dog has to be calm down immediately and has to be friendly towards the participating people and the judge. This kind of sharpness is defined as desired sharpness.
Reaction during blood drawn (RBD) The dog’s reactions while they are brought under treatment such as reactions after seeing needles during collection of blood from the animal’s body etc.
*

The traits were originally scored from high to low but were inverted for all traits to have a more intuitive score from low to high.

2

Adapted from Ruefenacht et al. (2002).

Table 2

Description of traits as quantitative and categorical measurements

Trait Original scoring Interpretation method* (Quantitative/categorical)
Nerve stability Very unstable, unstable, medium and stable Quantitative (as scored)
Binary (unstable, stable)
Affability Avoiding, less friendly, medium, friendly Quantitative (as scored)
Binary (non-friendly, friendly)
Wariness Very serious, serious, normal and absent Quantitative (as scored)
Binary (present, absent)
Adaptability Very weak, weak, medium and good Quantitative (as scored)
Binary (weak, good)
Activity Static, medium, active and, very active Quantitative (as scored)
Binary (non-active, active)
Sharpness Absent, normal and aggressive Quantitative (as scored)
Binary (absent, present)
Reaction during blood drawn Very agitated, shrunken, normal Quantitative (as scored)
Binary (abnormal, normal)
*

The analyses as both quantitative and categorical type were adapted from Han et al. (2010).

Table 3

Heritability estimates of behavior traits with quantitative values

Trait Heritability estimates

Univariate Multivariate


h2d h2m h2overall h2multi
Nerve stability 0.02 (0.04) 0.00 (0.00) 0.00 (0.03) 0.06 (0.00)
Affability 0.01 (0.04) 0.03 (0.03) 0.02 (0.03) 0.05 (0.00)
Wariness 0.04 (0.04) 0.00 (0.00) 0.05 (0.04) 0.09 (0.00)
Adaptability 0.07 (0.06) 0.03 (0.03) 0.09 (0.05) 0.11 (0.01)
Sharpness 0.09 (0.06) 0.02 (0.03) 0.11 (0.05) 0.10 (0.01)
Activity 0.16 (0.06) 0.01 (0.04) 0.16 (0.05) 0.17 (0.00)
Reaction during blood drawn 0.03 (0.05) 0.13 (0.04) 0.13 (0.05) 0.12 (0.01)

Table 4

Heritability estimates of behavior traits with binary values

Trait Heritability estimates

h2d h2m h2overall
Nerve stability 0.00 (0.00) 0.05 (0.10) 0.03 (0.12)
Affability 0.09 (0.16) 0.10 (0.14) 0.17 (0.14)
Wariness 0.02 (0.08) 0.00 (0.00) 0.02 (0.08)
Adaptability 0.00 (0.00) 0.13 (0.10) 0.11 (0.13)
Sharpness 0.06 (0.07) 0.06 (0.06) 0.11 (0.06)
Activity 0.10 (0.12) 0.00 (0.00) 0.10 (0.12)
Reaction during blood drawn 0.00 (0.00) 0.18 (0.05) 0.12 (0.06)

Table 5

Phenotypic (lower diagonal) and genetic correlation (upper diagonal) of behavior traits with quantitative values

Trait Nerve stability Affability Wariness Adaptability Sharpness Activity Reaction during blood drawn
Nerve stability - 0.94 0.86 0.90 0.69 0.75 −0.02
Affability 0.68 - 0.88 0.93 0.55 0.76 −0.04
Wariness 0.48 0.42 - 0.90 0.49 0.89 −0.25
Adaptability 0.60 0.62 0.46 - 0.46 0.74 0.05
Sharpness 0.14 0.16 −0.12 0.21 - 0.64 −0.38
Activity 0.30 0.31 0.24 0.39 0.26 - −0.52
Reaction during blood drawn 0.19 0.10 0.03 0.10 −0.07 −0.21 -

Table 6

Identities, positions, and effects of the SNPs associated with behavior traits in a Sapsaree population using Sap183 SNP panel

Trait/SNP markera SNPb CFA Kbpc Addd SEe −log10Pf Domg SE −log10P 2ph
Nerve stability 51.48
 BICF2P1415013 AG 1 53,277 −0.10 0.03 3.02 −0.07 0.04 3.02 4.66
 BICF2P312237 AC 9 41,129 0.10 0.03 4.15 - - - 8.64
 BICF2P290731 AG 12 45,632 - - - 0.18 0.04 3.58 5.26
 BICF2P866010 AC 14 50,483 0.53 0.09 6.42 −0.47 0.10 6.42 15.15
 BICF2S23618725 AC 18 36,864 0.09 0.03 3.41 −0.15 0.04 3.41 5.55
 BICF2P475745 AG 18 26,072 - - - −0.14 0.04 3.26 5.73
 BICF2S23138775 AG 33 23,576 −0.64 0.13 3.57 −0.54 0.14 3.57 6.50
Affability 66.75
 BICF2G630348856 AG 3 75,645 0.45 0.09 3.73 −0.37 0.10 3.73 5.02
 BICF2S23743883 AG 3 77,839 0.18 0.04 3.39 −0.15 0.05 3.39 3.23
 BICF2P76353 AG 8 39,540 −0.22 0.04 5.86 −0.21 0.05 5.86 7.03
 BICF2P720310 AG 10 64,228 0.32 0.05 8.46 −0.33 0.05 8.46 17.05
 BICF2S23748209 AG 18 23,025 0.21 0.04 4.64 −0.23 0.05 4.64 4.88
 BICF2S23660031 AG 24 30,243 −0.52 0.07 8.46 −0.54 0.08 8.46 13.24
 BICF2S23138775 AG 33 23,576 −0.79 0.09 6.99 −0.72 0.10 6.99 16.30
Wariness 46.53
 BICF2G630808634 AG 6 8,334 −0.11 0.02 4.93 - - - 9.27
 BICF2S2354669 AG 21 38,720 0.10 0.03 4.16 - - - 8.68
 TIGRP2P321376_rs9154926 AG 24 48,047 −0.36 0.06 4.70 0.49 0.09 4.70 11.28
 BICF2G630508563 AG 24 15,905 0.13 0.03 3.65 - - - 6.20
 BICF2S23341394 AG 27 43,064 −0.09 0.02 3.10 - - - 4.76
 BICF2G630274628 AG 28 15,800 0.08 0.02 3.20 - - - 6.35
Adaptability 44.37
 BICF2G630560546 AC 7 59,862 −0.72 0.13 5.21 −0.69 0.14 5.21 9.15
 BICF2G630202467 AG 17 51,813 −0.13 0.03 4.27 - - - 6.68
 BICF2P475745 AG 18 26,072 0.19 0.04 5.34 −0.23 0.06 5.34 12.78
 BICF2G630465505 AG 25 19,390 - - - 0.24 0.05 3.63 6.93
 BICF2G630622959 AG 29 15,035 0.35 0.10 4.13 −0.50 0.12 4.13 8.83
Sharpness 28.27
 BICF2S23610713 AC 10 52,206 - - - −0.23 0.06 3.39 6.97
 BICF2P593536 AG 12 32,489 - - - 0.24 0.06 3.11 6.05
 BICF2S23419449 AG 32 15,380 −0.10 0.04 3.13 −0.20 0.06 3.13 7.10
 BICF2G63068075 AG 38 24,667 - - - −0.30 0.06 3.57 8.15
Activity 40.28
 BICF2P426911 AC 1 56,681 −0.35 0.10 3.87 −0.48 0.11 3.87 9.44
 BICF2S23313765 AG 2 53,551 0.17 0.06 3.15 −0.24 0.08 3.15 6.56
 BICF2P930297 AG 12 51,575 - - - 0.25 0.07 3.01 4.99
 BICF2S23320857 AC 18 26,197 −0.33 0.09 3.21 −0.46 0.11 3.21 5.83
 BICF2P1154961 AG 32 13,634 - - - −0.23 0.07 3.35 7.13
 BICF2G630777247 AG 35 23,046 −0.55 0.15 3.12 −0.61 0.17 3.12 6.34
Reaction during blood drawn 46.04
 BICF2P205873 AC 7 80,452 0.14 0.04 5.29 - - - 12.65
 BICF2P443788 AG 17 18,839 −0.34 0.07 3.44 0.26 0.09 3.44 5.40
 BICF2S23327763 AG 18 19,132 −0.13 0.06 3.23 −0.27 0.08 3.23 6.24
 BICF2G630327289 AG 22 34,614 0.09 0.03 3.19 0.11 0.04 3.19 4.84
 BICF2G630470591 AG 25 9,620 −0.08 0.03 4.86 0.16 0.04 4.86 10.32
 BICF2G630805994 AG 26 13,567 0.19 0.04 3.99 −0.15 0.07 3.99 6.59

SNP, single nucleotide polymorphism; CFA, canine (canis familiaris) autosome; SE, standard error.

a,c

SNP marker annotations and their positions.

b

Nucleotides of substitution.

d

Estimate of additive effect is for allele substitution effect replacing the latter with the former allele (nucleotide) in the SNP column.

e

Standard error.

f

Negative logarithm of the comparison-wise p-value of the test statistic against the null hypothesis of no SNP effect at the SNP position.

g

Estimate is for allele dominance effect.

h

Estimates of genetic variance. Proportion of phenotypic variance explained by the SNP was mentioned in braces. The values on the rows of trait names are the sum of the %σ2p values across all the significant SNPs for the trait.