Multi-breed genome-wide association study reveals novel loci associated with the weight of internal organs
- Yuna He†1,
- Xinjian Li†1,
- Feng Zhang1,
- Ying Su1,
- Lijuan Hou1,
- Hao Chen1,
- Zhiyan Zhang1Email author and
- Lusheng Huang1Email author
© He et al. 2015
Received: 9 October 2014
Accepted: 30 October 2015
Published: 17 November 2015
Recently, many genome-wide association studies (GWAS) have been conducted to understand the genetic architecture of economic important traits in farm animals. Pig is widely used as a biomedical animal model for its similarity with humans in terms of organ formation and disease mechanisms. Moreover, understanding the mechanisms that underlie the development of internal organs will impact the productive potential of pigs. Our aim was to uncover new single nucleotide polymorphisms (SNPs) associated with the weight of internal organs and carcass and also potential candidate genes.
We performed GWAS for the weight of heart, liver, spleen, kidney and carcass on five pig populations (White Duroc × Erhualian F2 intercross, Sutai population, Laiwu population, Erhualian population and commercial population, for a total of 2650 individuals). Genotype data was produced using the PorcineSNP60 Beadchip array. After quality control, the data was used for association tests under a general linear mixed model. Population stratification was adjusted by including a random polygenic effect based on a matrix of genotypic relationships. A meta-analysis of our GWAS datasets was conducted by summing up the Chi square values across breeds, with the degrees of freedom of the Chi square distribution equal to the effective number of breeds.
Thirty-nine quantitative trait loci (QTL) located on 15 chromosomes were identified by the single-population GWAS at the suggestive level. Among these, nine QTL surpassed the 5 % genome-wide significance threshold, including four for heart weight on SSC (Sus scrofa chromosome) 2, 4, 7 and 10, two for liver weight on SSC7, two for spleen weight on SSC5 and SSC7 and one for carcass weight on SSC11. The QTL on SSC7 showed pleiotropic effects for heart, liver and spleen weights in the F2 population. In addition, two QTL were detected in several populations, including one on SSC2 for heart weight in the F2 and Sutai populations and one on SSC7 for liver weight in the F2 and Laiwu populations. The meta-analysis detected four novel QTL on SSC1, 3, 8 and 16 for carcass weight.
With the development of next-generation sequencing technologies [1, 2] and sophisticated statistical methods, it has become possible to identify the genetic basis of more and more economic important complex traits. For example, it has been shown that the genes insulin-like growth factor 2 (IGF2) and ryanodine receptor 1 (RYR1) are associated with an increase in muscle mass and a reduction in backfat thickness  and with pale, soft and exudative meat , respectively. Implementation of marker-assisted selection (MAS) in breeding schemes has greatly benefited the pig industry . In recent years, because of the advent of high-density single nucleotide polymorphisms (SNPs) genotyping panels, genome-wide association studies (GWAS) have been widely used to identify candidate genes and quantitative trait loci (QTL) for a number of complex traits in humans, animals and plants [6–9]. Compared to traditional QTL mapping strategies , GWAS do not require pedigree information or any prior assumption on the fixation of QTL alleles in the founder populations. These advantages make GWAS one of the most popular strategies for the study of complex diseases in humans or of economic important traits in farm animals .
Previous studies have shown that pigs and humans share similar characteristics in terms of body weight, physiological features, organ formation and mechanisms of disease infection . Using the pig as a biomedical model has led to significant progress in the study of atherosclerosis and diabetes . The pig is also considered as the most optimal donor of internal organs (e.g. heart) for humans [12, 14]. Deciphering the genetic architecture of complex traits such as the weight of internal organs in the pig will help to better understand the pathogenesis of human diseases associated with internal organs and to develop new gene therapy methods. Body weight is the sum of fat mass, internal organ mass, muscle mass and skeleton mass. Each of these components has its own developmental process and gene expression profile .
To date, the genetic architecture that underlies the weight of internal organs is not completely known. To our knowledge, most studies published on QTL associated with the weight of internal organs in pigs were based on QTL mapping strategies  and few used GWAS . Furthermore, most of these studies focused mainly on one breed or on intercross populations and, thus, the results only reflect a small proportion of the genetic mechanisms that underlie these complex traits and provide little information on the genetic homologies and differences between breeds. These limitations have hindered the progress in the fine-mapping and detection of additional QTL for these traits.
In this study, we conducted a GWAS on five pig populations that were genotyped with the PorcineSNP60 Beadchip array. Furthermore, a meta-analysis was performed by combining the five populations. We also attempted to highlight the genetic homologies and differences that underlie these complex quantitative traits between breeds and populations.
Animals and sample collection
Animal care and tissue collection procedures followed the guidelines established by the Ministry of Agriculture of China. The ethics committee of Jiangxi Agricultural University specifically approved this study.
Experimental animals were from five populations i.e.: (1) a White Duroc × Erhualian F2 intercross that derived from two White Duroc founder boars and 17 Chinese Erhualian founder sows (consisting of a sub-population of Chinese Taihu pigs) and comprised 1912 F2 pigs in six batches; (2) a Sutai population, which is a Chinese synthetic pig line that was originally generated from Chinese Taihu and Duroc pigs; the current population was generated by over 18 generations of artificial selection; in our study, we used 460 Sutai pigs from five sires and 60 dams; (3) a population of 316 Laiwu pigs; (4) a population of 334 Erhualian pigs that originated both from Chinese native purebred populations and from farms in Laiwu (Shandong province) and Changzhou (Jiangsu province), respectively; and (5) a commercial population that was a three-way-cross breed (Duroc × Large White × Yorkshire or DLY), for which 610 samples were collected. For the F2 and Sutai populations, fattening pigs were raised under consistent indoor conditions and were slaughtered at the age of 240 days. Only 1028 of the 1912 F2 individuals were slaughtered and recorded for phenotypic traits evaluated in this study; the other individuals were used for a study on male/female reproduction traits. Because the Chinese native breeds have a slower growth rate, they were fed ad libitum rather than on restricted feeding, which varied with age as for commercial breeds, and they were slaughtered at the older age of 300 days. DLY pigs were slaughtered at a weight of about 90 kg, which corresponds to about 180 days of age. Immediately after slaughter, we measured the weights of the heart, liver, spleen, kidney and carcass for the five populations except that kidney weight was not recorded for DLY pigs. Ear tissue was collected and stored in 75 % ethyl alcohol for DNA extraction.
Genotyping and quality control
For each animal, we extracted genomic DNA from ear tissue using a standard phenol/chloroform method. DNA was diluted to a final concentration of 50 ng/µL. Genotyping was performed using the PorcineSNP60 Beadchip array (Version 1 for the F2 and Sutai populations and Version 2 for the Laiwu and Erhualian pigs and the commercial population) according to the manufacturer’s protocol. Signal intensities from the Illumina iScan were normalized and genotypes were called using the Genome Studio software. Quality control procedures were carried out using PLINK v1.07 software , separately for each population. Briefly, samples with a call rate greater than 0.95 and a Mendelian error rate less than 0.05, and SNPs with a call rate greater than 0.95, a minor allele frequency (MAF) higher than 0.05 and a Mendelian error rate less than 0.05 were kept for further analysis. After quality control, 45,242, 46,347, 41,264, 30,410 and 51,668 SNPs and 928, 436, 317, 333 and 610 individuals remained for single-population GWAS for the F2, Sutai, Laiwu, Erhualian and DLY populations, respectively. In total, 45,962, 45,962, 45,962, 44,858 and 46,629 SNPs were retained for the meta-analysis that included heart weight, liver weight, spleen weight, kidney weight and carcass weight.
Correlations between traits and simple statistics were summarized using R software. Heritability was estimated using the polygenic function of the R package GenABEL . We then used the likelihood ratio test statistic (LRT) = −2ln(L0/L1) to test for the presence of heritable variance against the null hypothesis of no heritable variance, where L0 and L1 represented the likelihood values of the reduced (1) and full (2) models, respectively .
Single-population GWAS and meta-analysis
Association between each SNP and the phenotypic traits in each population was tested by fitting an additive model under a generalized linear mixed model using the GenABEL package , separately for each SNP and population. The model adjusted for population stratification by including a random polygenic effect, with the corresponding variance–covariance matrix proportionate to pairwise genome-wide identity-by-state. This model is denoted as:
For both single-population GWAS and the meta-analysis, suggestive and genome-wide significance thresholds  were determined by Bonferroni correction, which was defined as 1/N and 0.05/N, where N is the number of tested SNPs. Tested SNPs were positioned on pig chromosomes according to the current assembly of the pig genome (build 10.2) . A cluster of significant SNPs within a linkage disequilibrium (LD) block was treated as a single significant QTL. When the distance between two consecutive genome-wide significant SNPs was greater than 10 Mb, they were considered as representing two separate QTL .
The phenotypic variance explained by the top SNP was calculated with the R software as (Vreduce − Vfull)/Vphe, where Vfull and Vreduce are the residual variances of models with and without the SNP effect and Vphe is the phenotypic variance.
Descriptive statistics for the traits in pigs from five populations
Mean ± SD
0.34 ± 0.07
1.39 ± 0.27
0.13 ± 0.03
0.28 ± 0.06
34.65 ± 7.19
Mean ± SD
0.28 ± 0.06
1.11 ± 0.24
0.15 ± 0.04
0.25 ± 0.04
50.98 ± 13.01
Nb valid individuals
Mean ± SD
0.28 ± 0.04
1.16 ± 0.21
0.12 ± 0.05
0.23 ± 0.04
61.40 ± 12.05
Nb valid individuals
Mean ± SD
0.31 ± 0.05
1.24 ± 0.24
0.17 ± 0.04
0.27 ± 0.06
66.2 ± 11.40
Nb valid individuals
Mean ± SD
0.40 ± 0.05
1.68 ± 0.21
0.18 ± 0.04
84.95 ± 8.40
Genome-wide significant QTL identified by GWAS for the weight of four internal organs and carcass in five populations
Two genome-wide significant QTL and 12 suggestive QTL were identified for the Laiwu population (Table 2; Fig. 1). Of the two genome-wide significant QTL, a QTL for heart weight was located on SSC4, which included 20 SNPs, and a QTL for liver weight was located on SSC7, which included 24 SNPs. The QTL associated with liver weight on SSC7 comprised six significant SNPs that were shared between the F2 and Laiwu pig populations, which suggests that the variant responsible for the QTL effect on liver weight is the same in both populations.
For the other three populations (Erhualian, DLY commercial pigs, Sutai), several QTL were detected: (1) one genome-wide significant QTL at 97.9 Mb on SSC7 for spleen weight and three suggestive QTL, including one on SSC7 for heart weight and two on SSC7 and SSC12 for carcass weight in Erhualian pigs; six QTL in DLY commercial pigs, including two genome-wide significant QTL associated with spleen weight (SSC5 at 95.1 Mb) and heart weight (SSC10 at 32.2 Mb); and eight QTL in Sutai pigs, of which one achieved genome-wide significance level for carcass weight on SSC11 at 48.9 Mb (see Additional files 3, 4, 5, 6 and 7: Figures S1, S2, S3, S4, and S5).
Among the SNPs within these regions, the most significant SNP across all traits was rs80935535 (p = 1.91E−10) on SSC7 at 34.8 Mb, which accounted for 13.2 % of the phenotypic variance for liver weight in F2 pigs. This SNP was also associated with heart weight (p = 1.48E−09) and explained 13.2 % of the phenotypic variance. The second most significant SNP was rs81318741 (p = 2.0E−10) on SSC2 at 0.9 Mb, which explained 8.6 % of the phenotypic variance for heart weight in F2 pigs (Table 2; Fig. 1).
Meta-analysis of GWAS datasets
In this study, we detected 39 QTL for five traits in five experimental populations. Of these 39 QTL, nine reached the genome-wide significance level. Association signals were strongest for heart weight on SSC2, 4, 7 and for liver weight on SSC7 (Table 2; Fig. 1). Of these 39 QTL, eight confirmed previous reports from the pig QTL database (http://www.animalgenome.org/cgi-bin/QTLdb/SS/index). For example, the QTL for heart weight on SSC2 was identical to the QTL identified by Jeon et al.  and Wei et al.  and the QTL for liver heart on SSC7 was very similar to the QTL detected by Liu et al.  and Yue et al. .
These previous studies were mostly based on linkage analysis, which uses recent recombination events and thus, the mapping resolution is low (regions of more than 10 Mb and comprising hundreds of genes); in comparison, GWAS takes advantage of historical recombination events to reflect associations between markers and phenotypes, which results in a much higher resolution. Based on the rule of a LOD score cut-off of 2 , the confidence intervals of the regions that we identified on SSC2 in the F2 population were reduced to 0.3 Mb for heart weight (between 0.98 and 1.28 Mb) and 2.7 Mb for liver weight (between 34.7 and 37.4 Mb). Thus, GWAS is a powerful method to detect QTL that underlie complex traits or diseases and is currently applied for most major species.
Comparison of our results with those of previous QTL mapping studies
Compared to our previous QTL mapping study based on microsatellite markers in the F2 intercross population, which identified 22 significant QTL , this GWAS based on high-density SNPs revealed only nine significant QTL. The likely reason is that the sample size is smaller in our study (928 individuals instead of 1028 individuals used in Ma et al. ), which results in reduced power and in more stringent significance thresholds, since in the Ma et al. GWAS a larger number of tested SNPs was used for Bonferroni correction. Moreover, QTL linkage mapping assumes that QTL alleles are alternatively fixed in each founder breed of the F2 intercross. The GWAS model adopted in our study only included additive effects, while the QTL mapping strategy applied in  included both additive and dominance effects, which may be another reason why fewer QTL were detected in the GWAS analysis. To test the hypothesis that the current study may have missed QTL because of dominance, we performed a dominance GWAS using PLINK (v1.07) for kidney weight in the F2 population. We compared the significance of average phenotype differences between heterozygous and homozygous individuals at the tested SNP. The Manhattan plot of the dominance GWAS for kidney weight showed that among the three detected QTL, there were one genome-wide QTL and two suggestive QTL (see Additional file 9: Figure S6). Compared to the additive model that detected no QTL for kidney weight, the dominance model led to the identification of three novel QTL for this trait. Thus, some of the QTL that were missed in our GWAS were dominant.
The GWAS reported here confirmed two QTL that were identified in our previous linkage analysis  on the F2 population: one on SSC2 for heart weight and one on SSC4 for carcass weight; in addition, we detected four novel QTL, one for heart weight, one for liver weight and two for spleen weight. It should be noted that the QTL for liver weight on SSC7 was also identified by Liu et al.  within a 9.7 Mb region (between 31.24 and 41.00 Mb) but the GWAS allowed us to reduce the confidence interval to 2.7 Mb. This implies that linkage analysis and GWAS complement each other to detect genetic variants for traits of interest.
Genetic homologies and differences between populations
Using the F2 pig intercross, two major QTL were detected for heart weight on SSC2 at 0.9 Mb and on SSC7 at 34.7 Mb. These two QTL were also identified with the linkage mapping strategy . In contrast, no signal was detected for these traits when using the DLY commercial or Erhualian populations (that have founders similar to those of the F2 population), which implies that these two QTL were fixed in the founders of these two breeds or were segregating at a sufficiently low frequency such that they were not detected, or were false positives. Thus, this QTL could only be detected by using an intercross between these two breeds. Allele frequencies of the top SNP on SSC7 for heart weight were equal to 0.78 and 0.00 in the DLY and Erhualian breeds, respectively, which means that it was fixed in the Erhualian population. Of course, LD patterns between causative mutation and detected SNP can differ between populations and a much more complex LD structure may exist in the DLY hybrid population. Furthermore, in the Sutai population, a suggestive QTL for heart weight on SSC2 was replicated when analyzing the F2 population, which indicates that the QTL effect is homogeneous between the F2 and Sutai populations and thus that a common mutation is present. In the Laiwu population, a novel QTL was detected on SSC4 at 77.2 Mb but no association signal was identified for this region in any other population. Since this QTL was only observed in the Chinese specialized local breed, it indicates that the genetic architecture that underlies heart weight differs between the five populations studied here. Chinese native breeds cover about one third of the world’s genetic pig resources and since their genetic diversity is higher than that of commercial breeds, they increase the genetic variability of interbreed crosses.
Possible reasons for the differences observed between GWAS across populations  could be that: (a) the collective effect of many rare mutations plays a substantial role in complex phenotypes; (b) a common disease in unrelated affected individuals is caused by many (hundreds or even thousands) different rare severe mutations in the same gene; (c) the penetrance of a mutation differs between individuals; (d) mutations in different genes of the same or related pathway result in the same disorder; and (e) gene by gene and gene by environment interactions lead a causative mutation to have different genetic effects between populations . Therefore, understanding the genetic heterogeneity of traits will benefit personal medicine and therapy.
On SSC7, a pleiotropic QTL (top SNP rs80935535 at 34.8 Mb) was detected for both heart weight and liver weight in F2 pigs. Allele A inherited from the Erhualian breed increases the weights of liver and heart. Pleiotropy can be explained by different mechanisms: (1) alternative splicing events and alternate start or stop codons that can lead to different proteins and, thus, impact different phenotypes; (2) the gene network theory, which assumes that a key driver gene in a gene interaction network can impact several pathways and, thus, if this gene is mutated, pleiotropy is observed; and (3) a gene that has the same effect across multiple tissues . A highly significant correlation (r2 = [0.59 − 0.67]) between heart weight and liver weight was observed for the F2 pigs (see Additional file 1: Table S1), which suggests that the development of these two organs may be under the control of the same gene pathway. Previous studies showed that traits that are under the influence of pleiotropic QTL tend to cluster together or to be highly correlated, such as overall body size  and adiposity . The weight of organs can be expressed as a percentage of the body weight that varies with overall body size. Thus, the genetic mechanisms that are responsible for the weight of organs could be interrelated in some ways.
Plausible candidate genes in the identified QTL regions
SSC7 encompasses one major QTL for heart weight in the F2 population and for liver weight in the F2 and Laiwu populations. Within an interval of 5 Mb around the confidence interval ranging from 34.7 to 37.4 Mb, GRM4 (glutamate receptor, metabotropic 4) was identified as a strong candidate gene for this QTL based on its biological function and position, which was exactly at the same position as the most significant SNP (rs80935535, 34.8 Mb). GRM4 is expressed during the differentiation of embryonic stem cells into myocardial cells and its expression declines with maturation of the myocardial cells . HMGA1 (high mobility group AT-hook 1) is another candidate gene adjacent to GRM4 on SSC7 for heart weight and liver weight. HMGA1 has been reported to be associated with growth in pigs Liu et al. , and [36, 37] also considered this gene as the prime biological candidate for carcass traits and internal organ weights.
In conclusion, a total of 39 loci on 15 chromosomes were identified. The results confirmed some of the previously identified QTL but also identified several novel QTL for these traits. Besides, we identified three potential candidate genes, IGF2 for heart weight located on SSC2, GRM4 for heart weight and liver weight on SSC7 and HMGA1 for heart and liver weight on SSC7. We also compared genetic homologies and differences between these traits in different populations.
YH performed phenotype data collection and analysis, genotyping and drafted the manuscript. XL supervised the recording of phenotypic measures and performed statistical correction. FZ was responsible for the genotyping database and quality control. YS and HC participated in the collection of phenotype data. LH also participated in the collection of phenotype data and was responsible for DNA extraction. ZZ performed all GWAS analyses and helped to draft the manuscript. LH designed and supervised the project. All authors read and approved the final manuscript.
The authors acknowledge financial support from National Natural Science Foundation of China (31200926) and National 973 Program (2012CB722502).
The authors declare that they have no competing interests.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Metzker ML. Sequencing technologies—the next generation. Nat Rev Genet. 2009;11:31–46.View ArticlePubMedGoogle Scholar
- Mardis ER. The impact of next-generation sequencing technology on genetics. Trends Genet. 2008;24:133–41.View ArticlePubMedGoogle Scholar
- Van Laere AS, Nguyen M, Braunschweig M, Nezer C, Collette C, Moreau L, et al. A regulatory mutation in IGF2 causes a major QTL effect on muscle growth in the pig. Nature. 2003;425:832–6.View ArticlePubMedGoogle Scholar
- Fujii J, Otsu K, Zorzato F, de Leon S, Khanna VK, Weiler JE, et al. Identification of a mutation in porcine ryanodine receptor associated with malignant hyperthermia. Science. 1991;253:448–51.View ArticlePubMedGoogle Scholar
- Hayes B, Goddard ME. Evaluation of marker assisted selection in pig enterprises. Livest Prod Sci. 2003;81:197–211.View ArticleGoogle Scholar
- Xie L, Luo C, Zhang C, Zhang R, Tang J, Nie Q, et al. Genome-wide association study identified a narrow chromosome 1 region associated with chicken growth traits. PLoS One. 2012;7:e30910.PubMed CentralView ArticlePubMedGoogle Scholar
- Wang M, Yan J, Zhao J, Song W, Zhang X, Xiao Y, et al. Genome-wide association study (GWAS) of resistance to head smut in maize. Plant Sci (Amsterdam, Neth). 2012;196:125–31.Google Scholar
- Bolormaa S, Hayes B, Savin K, Hawken R, Barendse W, Arthur P, et al. Genome-wide association studies for feedlot and growth traits in cattle. J Anim Sci. 2011;89:1684–97.View ArticlePubMedGoogle Scholar
- Ren J, Mao H, Zhang Z, Xiao S, Ding N, Huang L. A 6-bp deletion in the TYRP1 gene causes the brown colouration phenotype in Chinese indigenous pigs. Heredity. 2010;106:862–8.PubMed CentralView ArticlePubMedGoogle Scholar
- Darvasi A. Experimental strategies for the genetic dissection of complex traits in animal models. Nat Genet. 1998;18:19–24.View ArticlePubMedGoogle Scholar
- Burton PR, Clayton DG, Cardon LR, Craddock N, Deloukas P, Duncanson A, et al. Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature. 2007;447:661–78.View ArticleGoogle Scholar
- Cooper D, Ye Y, Rolf L Jr, Zuhdi N. The pig as potential organ donor for man. In: Cooper DKE, Reemtsma K, White D, editors. Xenotransplantation. Berlin: Springer; 1991. p. 481–500.View ArticleGoogle Scholar
- Turk JR, Henderson KK, Vanvickle GD, Watkins J, Laughlin MH. Arterial endothelial function in a porcine model of early stage atherosclerotic vascular disease. Int J Exp Pathol. 2005;86:335–45.PubMed CentralView ArticlePubMedGoogle Scholar
- Ye Y, Niekrasz M, Kosanke S, Welsh R, Jordan H, Fox J, et al. The pig as a potential organ donor for man: a study of potentially transferable disease from donor pig to recipient man. Transplantation. 1994;57:694–702.View ArticlePubMedGoogle Scholar
- Zhang J, Xiong Y, Zuo B, Lei M, Jiang S, Fe Li, et al. Detection of quantitative trait loci associated with several internal organ traits and teat number trait in a pig population. J Genet Genomics. 2007;34:307–14.View ArticlePubMedGoogle Scholar
- Ma J, Ren J, Guo Y, Duan Y, Ding N, Zhou L, et al. Genome-wide identification of quantitative trait loci for carcass composition and meat quality in a large-scale White Duroc × Chinese Erhualian resource population. Anim Genet. 2009;40:637–47.View ArticlePubMedGoogle Scholar
- Liu X, Wang LG, Liang J, Yan H, Zhao KB, Li N, et al. Genome-Wide Association Study for certain carcass traits and organ weights in a Large White × Minzhu intercross porcine population. J Genet Genomics. 2014;13:2721–30.Google Scholar
- Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81:559–75.PubMed CentralView ArticlePubMedGoogle Scholar
- Aulchenko YS, Ripke S, Isaacs A, van Duijn CM. GenABEL: an R library for genome-wide association analysis. Bioinformatics. 2007;23:1294–6.View ArticlePubMedGoogle Scholar
- Bermingham ML, Bishop SC, Woolliams JA, Pong-Wong R, Allen AR, McBride SH, et al. Genome-wide association study identifies novel loci associated with resistance to bovine tuberculosis. Heredity (Edinb). 2014;112:543–51.PubMed CentralView ArticlePubMedGoogle Scholar
- Eding H, Meuwissen THE. Marker-based estimates of between and within population kinships for the conservation of genetic diversity. J Anim Breed Genet. 2001;118:141–59.View ArticleGoogle Scholar
- Fisher RA. Statistical methods for research workers. Biological monographs and manuals. Edinburgh: Oliver and Boyd; 1925.Google Scholar
- Yang Q, Cui J, Chazaro I, Cupples LA, Demissie S. Power and type I error rate of false discovery rate approaches in genome-wide association studies. BMC Genet. 2005;6(Suppl 1):S134.PubMed CentralView ArticlePubMedGoogle Scholar
- Cunningham F, Amode MR, Barrell D, Beal K, Billis K, Brent S, et al. Ensembl 2015. Nucleic Acids Res. 2015;43:D662–9.PubMed CentralView ArticlePubMedGoogle Scholar
- Zhang Z, Hong Y, Gao J, Xiao S, Ma J, Zhang W, et al. Genome-wide association study reveals constant and specific loci for hematological traits at three time stages in a White Duroc × Erhualian F2 resource population. PLoS One. 2013;8:e63665.PubMed CentralView ArticlePubMedGoogle Scholar
- Jeon JT, Carlborg Ö, Törnsten A, Giuffra E, Amarger V, Chardon P, et al. A paternally expressed QTL affecting skeletal and cardiac muscle mass in pigs maps to the IGF2 locus. Nat Genet. 1999;21:157–8.View ArticlePubMedGoogle Scholar
- Wei WH, Duan Y, Haley CS, Ren J, de Koning DJ, Huang LS. High throughput analyses of epistasis for swine body dimensions and organ weights. Anim Genet. 2011;42:15–21.View ArticlePubMedGoogle Scholar
- Yue G, Stratil A, Cepica S, Schröffel J, Schröffelova D, Fontanesi L, et al. Linkage and QTL mapping for Sus scrofa chromosome 7. J Anim Breed Genet. 2003;120:56–65.View ArticleGoogle Scholar
- Strachan T, Read AP. Genetic mapping of mendelian characters. In: Wiley L, editor. Human molecular genetics. New York: Garland Science; 1999.Google Scholar
- McClellan J, King MC. Genetic heterogeneity in human disease. Cell. 2010;141:210–7.View ArticlePubMedGoogle Scholar
- Hodgkin J. Seven types of pleiotropy. Int J Dev Biol. 1998;42:501–5.PubMedGoogle Scholar
- Leamy LJ, Pomp D, Eisen E, Cheverud JM. Pleiotropy of quantitative trait loci for organ weights and limb bone lengths in mice. Physiol Genomics. 2002;10:21–9.View ArticlePubMedGoogle Scholar
- Broman KW, Wu H, Sen Ś, Churchill GA. R/qtl: QTL mapping in experimental crosses. Bioinformatics. 2003;19:889–90.View ArticlePubMedGoogle Scholar
- Jungerius BJ, Van Laere A-S, Te Pas MFW, Van Oost BA, Andersson L, Groenen MAM. The IGF2-intron3-G3072A substitution explains a major imprinted QTL effect on backfat thickness in a Meishan × European white pig intercross. Genet Res. 2004;84:95–101.View ArticlePubMedGoogle Scholar
- Zheng B, Zhou LM, Zhao QW, Zhang XG, Zhu DY. Expression of metabotropic glutamate receptor 4 in cardiomyocytes differentiated from mouse embryonic stem cell in vitro. J Zhejiang Univ. 2012;41:366–72.Google Scholar
- Ai H, Ren J, Zhang Z, Ma J, Guo Y, Yang B, et al. Detection of quantitative trait loci for growth-and fatness-related traits in a large-scale White Duroc × Erhualian intercross pig population. Anim Genet. 2012;43:383–91.View ArticlePubMedGoogle Scholar
- Kim KS, Lee JJ, Shin HY, Choi BH, Lee CK, Kim JJ, et al. Association of melanocortin 4 receptor (MC4R) and high mobility group AT-hook 1 (HMGA1) polymorphisms with pig growth and fat deposition traits. Anim Genet. 2006;37:419–21.View ArticlePubMedGoogle Scholar