Multibreed genome wide association can improve precision of mapping causative variants underlying milk production in dairy cattle
© Raven et al.; licensee BioMed Central Ltd. 2014
Received: 17 May 2013
Accepted: 18 January 2014
Published: 24 January 2014
Genome wide association studies (GWAS) in most cattle breeds result in large genomic intervals of significant associations making it difficult to identify causal mutations. This is due to the extensive, low-level linkage disequilibrium within a cattle breed. As there is less linkage disequilibrium across breeds, multibreed GWAS may improve precision of causal variant mapping. Here we test this hypothesis in a Holstein and Jersey cattle data set with 17,925 individuals with records for production and functional traits and 632,003 SNP markers.
By using a cross validation strategy within the Holstein and Jersey data sets, we were able to identify and confirm a large number of QTL. As expected, the precision of mapping these QTL within the breeds was limited. In the multibreed analysis, we found that many loci were not segregating in both breeds. This was partly an artefact of power of the experiments, with the number of QTL shared between the breeds generally increasing with trait heritability. False discovery rates suggest that the multibreed analysis was less powerful than between breed analyses, in terms of how much genetic variance was explained by the detected QTL. However, the multibreed analysis could more accurately pinpoint the location of the well-described mutations affecting milk production such as DGAT1. Further, the significant SNP in the multibreed analysis were significantly enriched in genes regions, to a considerably greater extent than was observed in the single breed analyses. In addition, we have refined QTL on BTA5 and BTA19 to very small intervals and identified a small number of potential candidate genes in these, as well as in a number of other regions.
Where QTL are segregating across breed, multibreed GWAS can refine these to reasonably small genomic intervals. However, such QTL appear to represent only a fraction of the genetic variation. Our results suggest a significant proportion of QTL affecting milk production segregate within rather than across breeds, at least for Holstein and Jersey cattle.
Variation in milk production traits and functional traits in dairy cattle have a major genetic component . Genome wide association studies have been successful for identifying genomic regions which associate with these traits but few have led to identification of the underlying mutation (Eg. ). Within breed variation has been the focus of most analyses, however, within a cattle breed LD is extensive (eg r2 > 0.3 at 250 kb) making it difficult to map the mutation precisely, or even to a small number of candidate genes [3–7]. Between breeds however, linkage disequilibrium phase among SNP is only conserved at 5-10 kb, for Bos taurus breeds at least . So, expanding cattle GWAS to multiple breeds could potentially refine QTL intervals .
This strategy has been used extensively in dogs . The long distance LD in a single dog breed (r2 > 0.3 at 0.4-3.2 Mb) means that less markers are required for the initial association, but the precision of the mapping is poor . To overcome this problem a two-stage mapping strategy is applied. A single breed with long distance LD is used to identify QTL, then multiple breeds and dense SNP typing are used for fine mapping . As breeds can share ancestral mutations, recombination events can be used to refine QTL region . The multibreed strategy allows mapping to a region of ~100 kb (reflecting the ancestral haplotype block size in multibreed dog populations) which typically contains a limited number of candidate genes.
Applying a multibreed strategy in dairy cattle may be useful to refine the location of QTL. It should be remembered, however, that while modern cattle breeds have been created relatively recently (< 400 generations, ), in some cases there has been strong selection since breed divergence . As a result, some QTL segregating in one breed may not be segregating in the other. This study investigates the power of multibreed GWAS and explores the proportion of QTL segregating in multiple dairy cattle breeds, namely Holstein and Jersey dairy cattle. We hypothesise that a GWAS combining breeds will have more power to detect variation and that such GWAS should be able to map QTL to a smaller genomic interval. We conducted GWAS in 5240 Jerseys and 12685 Holsteins, both within breed and multibreed. Our results demonstrate that within breed analyses actually have more power to detect a higher proportion of the variation but multibreed GWAS results in the more precise mapping of the QTL that do segregate across breeds.
Within breed analysis
To firstly determine how well we could map and validate QTL within breeds, we split the Holstein and Jersey data sets into cows and bulls within each breed (Additional file 1: Table S1 describes number of phenotypes for each trait in each data set). A GWAS was conducted using a mixed model including a regression on SNP genotype (0, 1 or 2 copies of the second allele), fixed effects of breed and gender, and a random polygenic breeding value effect to account for population structure. The significant SNP (P < 10-8) in the two genders within a breed were then compared.
Results of within breed and within gender genome wide association analysis
Number of 100 kb intervals containing SNP significant in both breeds
Correlations of SNP effects for very significant SNP (P < 10 -20 ) within genders within each breed, and between breeds for milk production, reproduction and health traits
For some functional traits, namely fertility (measured as calving interval), there were no SNP significant at P < 10-8. Therefore, we also considered a suggestive significance threshold of P < 10-5 for these traits. In Holsteins, there were fewer SNP significant for SCC and Holstein cows displayed far fewer significant associations than Holstein bulls, which was also reflected in higher FDRs (Table 1; Additional file 3: Table S3). Jersey bulls showed a much larger number of significant SNP for the functional traits than Jersey cows (Table 1). Few SNP were significant for fertility and Jersey cows had many more significant SNP for SCC than Jersey bulls. There were few SNP shared within breeds for fertility and mammary system in Jerseys and fertility within Holsteins (Table 1). Although a much lower significance threshold was used, the proportion of common SNP for mammary system, survival and SCC still did not reach the levels observed for milk production traits.
The observation that fewer SNP are shared between cows and bulls for Jerseys than for Holsteins is likely a reflection of smaller sample size. Further, for non-production traits, the lower heritability of these traits likely leads to lower power, in both the Holstein and the Jersey analysis, and particularly for cows. There was a much larger number of cows in the analysis (Additional file 1: Table S1) but more accuracy associated with the phenotype for the bulls, which are daughter trait deviations from thousands of daughters in some cases. There were especially large variations in the number of significant SNP between genders for mammary system and survival.
Comparison of QTL between breeds
Results of within breed genome wide association analysis
Number of 100 kbintervals containing SNP significant in both breeds
Fewer significant associations were found for health and reproductive traits in either breed. However, for mammary system, a number of significant associations were detected in Holsteins and also for survival in Jerseys (Table 3). Health and reproductive traits were significant at P < 10-8 across many chromosomes, however at P < 10-20 we were only able to identify major QTL for BTA18 for survival, BTA11 for mammary system and BTA10 for SCC in Holsteins. Jerseys showed very significant QTL on BTA2, BTA6, BTA11 and BTA25 for survival. False discovery rates were very high for fertility, particularly in Jerseys (Additional file 3: Table S3). Estimates for survival and SCC appear less powerful than milk production traits with moderate error rates, again likely reflecting lower heritabilities for these traits. When the suggestive threshold was used (P < 10-5) there were still few SNP common between Holsteins and Jerseys for these traits. Where significant SNP were common to each breed, we used a correlation analysis to determine whether the direction of effects was the same in both breeds. We identified a strong positive correlation between the common SNP significantly associating with survival, however, there were few significant SNP in the comparison, so sample size may be inadequate to affirm this (Table 2).
Most significant SNP regions and candidate genes for production and functional traits in dairy cattle from a large multibreed GWAS
100 kb LD
<500 kb LD
diacylglycerol O-acyltransferase 1
Beta-transducin repeat containing
microsomal glutathione S-transferase 1
zinc finger protein 85-like
ring finger protein 219
diacylglycerol O-acyltransferase 1
epidermal growth factor receptor pathway substrate 8
group-specific component (vitamin D binding protein)
RNA polymerase II associated protein 3
diacylglycerol O-acyltransferase 1
phospholipase C, epsilon 1
phosphatidic acid phosphatase type 2 domain containing 1A
diacylglycerol O-acyltransferase 1
microsomal glutathione S-transferase 1
GINS complex subunit 4 (Sld5 homolog)
SP110 nuclear body protein
diacylglycerol O-acyltransferase 1
casein alpha s1
glucosidase, beta, acid
RNA polymerase II associated protein 3
carcinoembryonic antigen-related cell adhesion molecule 18-like
neuropeptide FF receptor 2
SRY (sex determining region Y)-box 5
HIG1 domain family, member 1D-like pseudogene
magnesium transporter NIPA2-like
suppressor of fused homolog (Drosophila)
SWI/SNF related, matrix assoc., actin dep. Reg. of chromatin, subfamily a-like 1
protein kinase C, epsilon
ISL LIM homeobox 1
NLR family, pyrin domain containing 6
WD repeat domain 27
transducin (beta)-like 1X-linked
mesoderm induction early response 1, family member 3-like
neuropeptide Y receptor Y2
MIS18 kinetochore protein homolog A (S. pombe)
CTTNBP2 N-terminal like
La ribonucleoprotein domain family, member 1
aarF domain containing kinase 3
Comparison of the proportion of SNP occurring within genes in the whole genome (632 K SNP) and SNP significant (P < 10 -8 ) in the multibreed and single breed analyses
Proportion of SNP Within Genes
632 K SNP
Multibreed (P < 10-8)
Holsteins (P < 10-8)
Jerseys (P < 10-8)
We performed a GWAS in a large multibreed dairy cow population with 632,003 SNP to identify the genomic regions associating with milk production, health and reproduction traits. We aimed to determine whether multibreed data sets are more powerful and result in more precise mapping of QTL. The results suggest that the multibreed analysis is actually less powerful in terms of number of QTLs identified, as there are a considerable number of QTL that only segregate in one breed. However, when QTL did segregate across breeds, the multibreed analysis refined QTL to smaller genomic regions, allowing a small number of potential candidates to be identified.
Results for functional traits were inconsistent within and between breeds. For example, there was a limited overlap of significant QTL within the Jersey breed. Lower heritabilities (in particular larger environmental variation) reduced power for these traits particularly in Jerseys and in cows of both breeds. We observed very few QTL that segregated across breeds for these traits. Again, this is likely an artefact of the low power for these traits. Another possible explanation would be strong selection for these traits has fixed QTL in one breed and not the other. The common significant SNP shared within Holsteins (between bulls and cows) were highly positively correlated, in effect and in direction. This suggests that sample size may be more important than gender differences in power. This was previously shown in beef breeds where using breeds with smaller sample sizes reduced the power to identify segregating loci .
Very significant QTL (P < 10-20) were identified in both the multibreed analysis, and the within breed analyses. Among these were several QTL that had been previously described. SNP close to DGAT1 were highly associated for all milk production traits in both breeds. A very significant QTL on BTA20 has been previously ascribed to a mutation in GHR[15, 19], but is more likely a number of separate QTL given the results of our multibreed analysis. Another very significant QTL was identified on BTA14 at approximately 70Mb for fat percentage, protein percentage and milk. This QTL appears to locate around the SDC2 gene. Previous studies have shown evidence of QTL at the telomeric end of BTA14 but this was not confirmed here . We also identified a major QTL on BTA6 for protein kg, which centres within CSN1. For protein kg, a major QTL on BTA5 centres within CD27.
The value of the multibreed analysis was highlighted by the fact that more SNP were located within genes than when breeds were separated, suggesting there is more precision in refining QTL in the multibreed analysis. Inspection of several QTL regions support this. For example, a region on BTA19 was refined such that a strong candidate, the lipid metabolism gene ACLY, could be identified. ACLY is differentially express in rat mammary tissue during pregnancy and lactation . This region of BTA19 was identified in a GWAS phenotyping the ratio of saturated to unsaturated fats in the milk of Danish Holstein cows but is otherwise undescribed . The multibreed analysis shows a clear benefit over breed dependent methods in refining the precise locations of these fat metabolism genes.
The very significant QTL affecting fat kg and fat percentage on BTA5 centred within 3000 bp of MGST1, an inflammation response gene which is highly expressed through pregnancy and lactation . MGST1 was upregulated during adipocyte development in the Longissimus muscle in Japanese Black cattle . Slightly upstream, a highly significant peak centred within EPS8. EPS8 acts as a receptor tyrosine kinase substrate for epidermal growth factor receptor (EGFR) and thus increases the signalling response to epidermal growth factor (EGF) . A previous study on German Holsteins identified EPS8 as a candidate fat percentage gene . Another study found a milk yield QTL from 92.1-93 Mb but did not localise to near MGST1. Our results suggest both MGST1 and EPS8 may contain QTL affecting milk production.
Other QTL have been identified at the telomeric end of this chromosome which do not align with our results. Cole et al. identified fat yield QTL at 98.7 Mb, but our peak was 1 Mb downstream within LOH12CR1. A 50 k GWAS of Canadian Holsteins identified several fat yield candidate genes including SLC2A3 and LOC786521 (GDF3) at 101.7-101.8 Mb and LRP6, LOC786490 (EMP1) and DUSP16 at 97-98 Mb . This region was highly significant for milk and protein in our study rather than fat and their candidate genes do not lie within our QTL peak.
A limitation of this study is the use of a set of ascertained SNP to map QTL. Interpretations regarding the position of QTL which segregate across breeds must be treated with caution, as this will depend on the extent of linkage disequilibrium between breeds with causative mutations. As whole genome sequence data becomes increasingly available, this study could be replicated using imputed, full-sequence genotypes to determine whether the Holsteins and Jerseys still maintain a larger proportion of significant variants than the multibreed sample.
Finally, population stratification is a key cause of false positive results in GWAS as admixture occurs subtly in the form of relationships among animals . In our analysis, where breeds are combined, we have attempted to account for stratification both by fitting a breed effect and pedigree effect within breed. As previously seen (and in QQ plots, Additional file 5: Figure S1), our within breed analysis actually reduced the number of significant effects suggesting combining the breeds is not leading to an increase in the number of false positives [6, 28]. Finally, multibreed GWAS is likely to identify older, conserved mutations but may not be as effective as a single breed model for recently diverged mutations.
A multibreed analysis together with dense SNP genotypes has allowed us to refine QTL locations for milk production and functional traits – for example this approach allowed us to refined QTL on BTA5 and BTA19 to a limited number of candidate genes. Further evidence that the multibreed analysis refines QTL regions is that we observed an enrichment of significant SNP within genes in the multibreed analysis. However, there is still a considerable role for studies on individual breeds as our results suggest a considerable proportion of QTL do not segregate across breeds (for Holstein and Jersey cattle at least). In future, using sequence data rather than SNP array genotypes combined with a multibreed analysis could potentially lead to direct identification of the causative mutation.
Phenotypes and genotypes
There were 17925 dairy cattle in the study in total. Of these 9289 were Holstein cows, 3396 were Holstein bulls, 4226 were Jersey Cows and 1014 were Jersey Bulls (Additional file 1: Table S1). Phenotype records were available for a range of milk production, health and reproductive traits including milk yield, fat yield, protein yield, fat%, protein%, protein kg, fertility (calving interval), mammary system, somatic cell count and survival [29, 30]. The phenotypes were trait deviations for cows and daughter yield deviations for bulls. Trait deviations were calculated from raw phenotypes in a very large data set of approximately one million cows, fitting a model which included herd year season, age of cow and permanent environment effect. Daughter trait deviations for bulls were then calculated from the trait deviations of their daughters, corrected for breed of mate. Records were standardised to have a mean of zero and standard deviation of one in both breeds.
Each animal was SNP genotyped either using the Illumina BovineHD BeadChip or the Illumina 50 k Bovine chip. Filtering of SNP was performed as described by . Animals genotyped for 50 k were imputed to HD genotypes as described by Erbe, Hayes et al. (2012). Map positions of these SNP (and candidate genes) were from the Bovine Genome UMD3.1 assembly in the NCBI database (NCBI, http://www.ncbi.nlm.nih.gov).
Animals were divided into subsets for analysis (Additional file 1: Table S1). First, the sample was divided by gender within breed to create a dataset for within breed comparisons. To compare the proportion of significant SNP for each trait for each breed, cows and bulls within each breed were analysed together. Finally, the entire sample of 17925 animals was used for a full multibreed analysis (Additional file 1: Table S1).
where, m is the number of tests, P is the probability value of the F-test and S is the proportion of significant SNP . A Q-Q plot showing the distribution of significant effects is provided (Additional file 5: Figure S1).
For Holstein cows and Holstein bulls and Jersey cows and Jersey bulls we assessed how many of the significant SNP were the same in both GWAS. We then expanded to 100 kb either side of the SNP in order to account for the possibility that the same causative mutation may be in LD with a different SNP in the different breeds. We also correlated the effects of these SNP for the two breeds to determine if the SNP were in the same linkage phase with the QTL. We also investigated genes reported to have mutations affecting milk production, namely DGAT1, GHR and ABCG2 , to further describe the precision of mapping within breeds and in the multibreed analysis for refining QTL position [14, 19, 37].
Availability and requirements
A full SNP map and P-values for 9 traits are provided in Additional file 6. Programs, scripts and information for setting up the analysis can be obtained from the authors upon request.
There were no animal studies conducted for this manuscript. Where animal data was used references have been provided.
We would like to thank the Dairy Futures CRC for funding and their support for the project.
- Cole JB, VanRaden PM, O’Connell JR, Van Tassell CP, Sonstegard TS, Schnabel RD, Taylor JF, Wiggans GR: Distribution and location of genetic effects for dairy traits. J Dairy Sci. 2009, 92 (6): 2931-2946. 10.3168/jds.2008-1762.PubMedView ArticleGoogle Scholar
- Cole J, Wiggans G, Ma L, Sonstegard T, Lawlor T, Crooker B, Van Tassell C, Yang J, Wang S, Matukumalli L, et al: Genome-wide association analysis of thirty one production, health, reproduction and body conformation traits in contemporary U.S. Holstein cows. BMC Gen. 2011, 12 (1): 408-425. 10.1186/1471-2164-12-408.View ArticleGoogle Scholar
- Melka M, Schenkel F: Analysis of genetic diversity in Brown Swiss, Jersey and Holstein populations using genome-wide single nucleotide polymorphism markers. BMC Res Notes. 2012, 5 (1): 161-168. 10.1186/1756-0500-5-161.PubMed CentralPubMedView ArticleGoogle Scholar
- Olson KM, VanRaden PM, Tooker ME: Multibreed genomic evaluations using purebred Holsteins, Jerseys, and Brown Swiss. J Dairy Sci. 2012, 95 (9): 5378-5383. 10.3168/jds.2011-5006.PubMedView ArticleGoogle Scholar
- Pant SD, Schenkel FS, Verschoor CP, Karrow NA: Use of breed-specific single nucleotide polymorphisms to discriminate between Holstein and Jersey dairy cattle breeds. Anim Biotechnol. 2012, 23 (1): 1-10. 10.1080/10495398.2012.636224.PubMedView ArticleGoogle Scholar
- VanRaden PM, Tooker ME, Cole JB, Wiggans GR, Megonigal JH: Genetic evaluations for mixed-breed populations. J D Sci. 2007, 90 (5): 2434-2441. 10.3168/jds.2006-704.View ArticleGoogle Scholar
- The Bovine Hap Map Consortium: Genome-wide survey of SNP variation uncovers the genetic structure of cattle breeds. Science (New York, NY). 2009, 324 (5926): 528-532.View ArticleGoogle Scholar
- Goddard ME, Hayes BJ: Mapping genes for complex traits in domestic animals and their use in breeding programmes. Nat Rev Genet. 2009, 10 (6): 381-391. 10.1038/nrg2575.PubMedView ArticleGoogle Scholar
- Sutter NB, Eberle MA, Parker HG, Pullar BJ, Kirkness EF, Kruglyak L, Ostrander EA: Extensive and breed-specific linkage disequilibrium in Canis familiaris. Genome Res. 2004, 14 (12): 2388-2396. 10.1101/gr.3147604.PubMed CentralPubMedView ArticleGoogle Scholar
- Karlsson EK, Baranowska I, Wade CM, Salmon Hillbertz NHC, Zody MC, Anderson N, Biagi TM, Patterson N, Pielberg GR, Kulbokas EJ, et al: Efficient mapping of mendelian traits in dogs through genome-wide association. Nat Genet. 2007, 39 (11): 1321-1328. 10.1038/ng.2007.10.PubMedView ArticleGoogle Scholar
- Ostrander EA, Kruglyak L: Unleashing the canine genome. Genome Res. 2000, 10 (9): 1271-1274. 10.1101/gr.155900.PubMedView ArticleGoogle Scholar
- de Roos APW, Hayes BJ, Spelman RJ, Goddard ME: Linkage disequilibrium and persistence of phase in Holstein–Friesian, Jersey and Angus Cattle. Genetics. 2008, 179 (3): 1503-1512. 10.1534/genetics.107.084301.PubMed CentralPubMedView ArticleGoogle Scholar
- Willis MB: The genetics of cattle. Heredity. 2000, 84 (1): 131-132.View ArticleGoogle Scholar
- Cohen-Zinder M, Seroussi E, Larkin DM, Loor JJ, Wind AE-v, Lee J-H, Drackley JK, Band MR, Hernandez AG, Shani M, et al: Identification of a missense mutation in the bovine ABCG2 gene with a major effect on the QTL on chromosome 6 affecting milk yield and composition in Holstein cattle. Genome Res. 2005, 15 (7): 936-944. 10.1101/gr.3806705.PubMed CentralPubMedView ArticleGoogle Scholar
- Chamberlain AJ, Hayes BJ, Savin K, Bolormaa S, McPartlan HC, Bowman PJ, Van Der Jagt C, MacEachern S, Goddard ME: Validation of single nucleotide polymorphisms associated with milk production traits in dairy cattle. J Dairy Sci. 2012, 95 (2): 864-875. 10.3168/jds.2010-3786.PubMedView ArticleGoogle Scholar
- van Dorland HA, Richter S, Morel I, Doherr MG, Castro N, Bruckmaier RM: Variation in hepatic regulation of metabolism during the dry period and in early lactation in dairy cows. J Dairy Sci. 2009, 92 (5): 1924-1940. 10.3168/jds.2008-1454.PubMedView ArticleGoogle Scholar
- Bouwman A, Bovenhuis H, Visker M, van Arendonk J: Genome-wide association of milk fatty acids in Dutch dairy cattle. BMC Genet. 2011, 12 (1): 43-PubMed CentralPubMedView ArticleGoogle Scholar
- McClure MC, Ramey HR, Rolf MM, McKay SD, Decker JE, Chapple RH, Kim JW, Taxis TM, Weaber RL, Schnabel RD, et al: Genome-wide association analysis for quantitative trait loci influencing Warner–Bratzler shear force in five taurine cattle breeds. Anim Genet. 2012, 43 (6): 662-673. 10.1111/j.1365-2052.2012.02323.x.PubMed CentralPubMedView ArticleGoogle Scholar
- Blott S, Kim J-J, Moisio S, Schmidt-Kuntzel A, Cornet A, Berzi P, Cambisano N, Ford C, Grisart B, Johnson D, et al: Molecular dissection of a quantitative trait locus: a phenylalanine-to-tyrosine substitution in the transmembrane domain of the bovine growth hormone receptor is associated with a major effect on milk yield and composition. Genetics. 2003, 163 (1): 253-266.PubMed CentralPubMedGoogle Scholar
- Heyen DW, Weller JI, Ron M, Band M, Beever JE, Feldmesser E, Da Y, Wiggans GR, VanRaden PM, Lewin HA: A genome scan for QTL influencing milk production and health traits in dairy cattle. Physiol Genomics. 1999, 1 (3): 165-175.PubMedGoogle Scholar
- Casey T, Zakrzewska EI, Maple RL, Lintault L, Wade CE, Baer LA, Ronca AE, Plaut K: Hypergravity disruption of homeorhetic adaptations to lactation in rat dams include changes in circadian clocks. Biology Open. 2012, 1 (6): 570-581. 10.1242/bio.2012687.PubMed CentralPubMedView ArticleGoogle Scholar
- Church PC, Goscinski A, Lefevre C: EXP-PAC: Providing comparative analysis and storage of next generation gene expression data. Genomics. 2012, 100 (1): 8-13. 10.1016/j.ygeno.2012.05.007.PubMedView ArticleGoogle Scholar
- Wang YH, Reverter A, Mannen H, Taniguchi M, Harper GS, Oyama K, Byrne KA, Oka A, Tsuji S, Lehnert SA: Transcriptional profiling of muscle tissue in growing Japanese Black cattle to identify genes involved with the development of intramuscular fat. Aust J Exp Agric. 2005, 45 (8): 809-820. 10.1071/EA05058.View ArticleGoogle Scholar
- Fazioli F, Minichiello L, Matoska V, Castagnino P, Miki T, Wong WT, Di Fiore PP: Eps8, a substrate for the epidermal growth factor receptor kinase, enhances EGF-dependent mitogenic signals. EMBO J. 1993, 12 (10): 3799-3808.PubMed CentralPubMedGoogle Scholar
- Wang X, Wurmser C, Pausch H, Jung S, Reinhardt F, Tetens J, Thaller G, Fries R: Identification and dissection of four major QTL affecting milk fat content in the German Holstein-Friesian population. PLoS ONE. 2012, 7 (7): e40711-10.1371/journal.pone.0040711.PubMed CentralPubMedView ArticleGoogle Scholar
- Mai MD, Sahana G, Christiansen FB, Guldbrandtsen B: A genome-wide association study for milk production traits in Danish Jersey cattle using a 50K single nucleotide polymorphism chip. J Anim Sci. 2010, 88 (11): 3522-3528. 10.2527/jas.2009-2713.PubMedView ArticleGoogle Scholar
- Li H, Wang Z, Moore SS, Schenkel FS, Stothard P: Genome-wide scan for positional and functional candidate genes affecting milk production traits in Canadian Holstein Cattle. Proc 9th WCGALP, Leipzig, Germany. 2010, 26:Google Scholar
- Ma L, Wiggans G, Wang S, Sonstegard T, Yang J, Crooker B, Cole J, Van Tassell C, Lawlor T, Da Y: Effect of sample stratification on dairy GWAS results. BMC Genomics. 2012, 13 (1): 536-10.1186/1471-2164-13-536.PubMed CentralPubMedView ArticleGoogle Scholar
- Moser G, Khatkar M, Hayes B, Raadsma H: Accuracy of direct genomic values in Holstein bulls and cows using subsets of SNP markers. Genet Sel Evol. 2010, 42 (1): 1-15. 10.1186/1297-9686-42-1.View ArticleGoogle Scholar
- Pryce JE, Bolormaa S, Chamberlain AJ, Bowman PJ, Savin K, Goddard ME, Hayes BJ: A validated genome-wide association study in 2 dairy cattle breeds for milk production and fertility traits using variable length haplotypes. J Dairy Sci. 2010, 93 (7): 3331-3345. 10.3168/jds.2009-2893.PubMedView ArticleGoogle Scholar
- Erbe M, Hayes BJ, Matukumalli LK, Goswami S, Bowman PJ, Reich CM, Mason BA, Goddard ME: Improving accuracy of genomic predictions within and between dairy cattle breeds with imputed high-density single nucleotide polymorphism panels. J Dairy Sci. 2012, 95 (7): 4114-4129. 10.3168/jds.2011-5019.PubMedView ArticleGoogle Scholar
- NCBI National Centre for Biotechnology Information.http://www.ncbi.nlm.nih.gov/,
- Gilmour AR, Gogel BJ, Cullis BR, Welham SJ, Thompson R: ASReml User Guide Release 1.0. VSN International Ltd. 2002, UK: Hemel Hempstead, HP1, 1ESGoogle Scholar
- Garrick DJ, Taylor JF, Fernando RL: Deregressing estimated breeding values and weighting information for genomic regression analyses. Genet Sel Evol. 2009, 41: 55-10.1186/1297-9686-41-55.PubMed CentralPubMedView ArticleGoogle Scholar
- MacLeod IM, Hayes BJ, Savin KW, Chamberlain AJ, McPartlan HC, Goddard ME: Power of a genome scan to detect and locate quantitative trait loci in cattle using dense single nucleotide polymorphisms. J Anim Breed Genet. 2010, 127 (2): 133-142. 10.1111/j.1439-0388.2009.00831.x.PubMedView ArticleGoogle Scholar
- Bolormaa S, Pryce JE, Hayes BJ, Goddard ME: Multivariate analysis of a genome-wide association study in dairy cattle. J Dairy Sci. 2010, 93 (8): 3818-3833. 10.3168/jds.2009-2980.PubMedView ArticleGoogle Scholar
- Grisart B, Coppieters W, Farnir F, Karim L, Ford C, Berzi P, Cambisano N, Mni M, Reid S, Simon P, et al: Positional candidate cloning of a QTL in dairy cattle: identification of a missense mutation in the bovine DGAT1 gene with major effect on milk yield and composition. Genome Res. 2002, 12 (2): 222-231. 10.1101/gr.224202.PubMedView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.