Genome‐wide association study for body weight in cattle populations from Siberia

Body weight is a complex trait in cattle associated with commonly used commercial breeding measurements related to growth. Although many quantitative trait loci (QTL) for body weight have been identified in cattle so far, searching for genetic determinants in different breeds or environments is promising. Therefore, we carried out a genome‐wide association study (GWAS) in two cattle populations from the Russian Federation (Siberian region) using the GGP HD150K array containing 139 376 single nucleotide polymorphism (SNP) markers. Association tests for 107 550 SNPs left after filtering revealed five statistically significant SNPs on BTA5, considering a false discovery rate of less than 0.05. The chromosomal region containing these five SNPs contains the CCND2 gene, which was previously associated with average daily weight gain and body mass index in US beef cattle populations and in humans respectively. Our study is the first GWAS for body weight in beef cattle populations from the Russian Federation. The results provided here suggest that, despite the existence of breed‐ and species‐specific QTL, the genetic architecture of body weight could be evolutionarily conserved in mammals.     

A post‐GWAS confirming the SCD gene associated with milk medium‐ and long‐chain unsaturated fatty acids in Chinese Holstein population

The stearoyl‐CoA desaturase (delta‐9‐desaturase) gene encodes a key enzyme in the cellular biosynthesis of monounsaturated fatty acids. In our initial genome‐wide association study (GWAS) of Chinese Holstein cows, 19 SNPs fell in a 1.8‐Mb region (20.3–22.1 Mb) on chromosome 26 underlying the SCD gene and were highly significantly associated with C14:1 or C14 index. The aims of this study were to verify whether the SCD gene has significant genetic effects on milk fatty acid composition in dairy cattle. By resequencing the entire coding region of the bovine SCD gene, a total of six variations were identified, including three coding variations (g.10153G>A, g.10213T>C and g.10329C>T) and three intronic variations (g.6926A>G, g.8646G>A and g.16158G>C). The SNP in exon 3, g.10329C>T, was predicted to result in an amino acid replacement from alanine (GCG) to valine (GTG) in the SCD protein. An association study for 16 milk fatty acids using 346 Chinese Holstein cows with accurate phenotypes and genotypes was performed using the mixed animal model with the proc mixed procedure in sas 9.2. All six detected SNPs were revealed to be associated with six medium‐ and long‐chain unsaturated fatty acids (P = 0.0457 to P < 0.0001), specifically for C14:1 and C14 index (P = 0.0005 to P < 0.0001). Subsequently, strong linkage disequilibrium (D′ = 0.88–1.00) was observed among all six SNPs in SCD and the five SNPs (rs41623887, rs109923480, rs42090224, rs42092174 and rs42091426) within the 1.8‐Mb region identified in our previous GWAS, indicating that the significant association of the SCD gene with milk fatty acid content traits reduced the observed significant 1.8‐Mb chromosome region in GWAS. Haplotype‐based analysis revealed significant associations of the haplotypes encompassing the six SCD SNPs and one SNP (rs109923480) in a GWAS with C14:1, C14 index, C16:1 and C16 index (P = 0.0011 to P < 0.0001). In summary, our findings provide replicate evidence for our previous GWAS and demonstrate that variants in the SCD gene are significantly associated with milk fatty acid composition in dairy cattle, which provides clear evidence for an increased understanding of milk fatty acid synthesis and enhances opportunities to improve milk‐fat composition in dairy cattle.     

Genome‐wide association study uncovers four QTL predisposing to supernumerary teats in cattle

Supernumerary teats (hyperthelia, SNTs) are a common abnormality of the bovine udder with a medium to high heritability and a postulated oligogenic or polygenic inheritance pattern. SNTs not only negatively affect machine milking ability but also act as a reservoir for bacteria. A genome‐wide association study was carried out to identify genes involved in the development of SNTs in the dual‐purpose Fleckvieh breed. A total of 2467 progeny‐tested bulls were genotyped at 43 698 single nucleotide polymorphisms, and daughter yield deviations (DYDs) for ‘udder clearness' (UC) were used as high‐heritability phenotypes. Massive structuring of the study population was accounted for by principal components analysis‐based and mixed model‐based approaches. Four loci on BTA5, BTA6, BTA11 and BTA17 were significantly associated with the UC DYD. Three associated regions contain genes of the highly conserved Wnt signalling pathway. The four QTL together account for 10.7% of the variance of the UC DYD, whereas the major fraction of the DYD variance is attributable to chromosomes with no identified QTL. Our results support both an oligogenic and a polygenic inheritance pattern of SNTs in cattle. The identified candidate genes permit insights into the genetic architecture of teat malformations in cattle and provide clues to unravel the molecular mechanisms of mammary gland alterations in cattle and other species.     

Identification of loci governing eight agronomic traits using a GBS‐GWAS approach and validation by QTL mapping in soya bean

Soya bean is a major source of edible oil and protein for human consumption as well as animal feed. Understanding the genetic basis of different traits in soya bean will provide important insights for improving breeding strategies for this crop. A genome‐wide association study (GWAS) was conducted to accelerate molecular breeding for the improvement of agronomic traits in soya bean. A genotyping‐by‐sequencing (GBS) approach was used to provide dense genome‐wide marker coverage (>47 000 SNPs) for a panel of 304 short‐season soya bean lines. A subset of 139 lines, representative of the diversity among these, was characterized phenotypically for eight traits under six environments (3 sites × 2 years). Marker coverage proved sufficient to ensure highly significant associations between the genes known to control simple traits (flower, hilum and pubescence colour) and flanking SNPs. Between one and eight genomic loci associated with more complex traits (maturity, plant height, seed weight, seed oil and protein) were also identified. Importantly, most of these GWAS loci were located within genomic regions identified by previously reported quantitative trait locus (QTL) for these traits. In some cases, the reported QTLs were also successfully validated by additional QTL mapping in a biparental population. This study demonstrates that integrating GBS and GWAS can be used as a powerful complementary approach to classical biparental mapping for dissecting complex traits in soya bean.     

Three novel quantitative trait loci for skin thickness in swine identified by linkage and genome‐wide association studies

Skin is the largest organ in the pig body and plays a key role in protecting the body against pathogens and excessive water loss. Deciphering the genetic basis of swine skin thickness would enrich our knowledge about the skin. To identify the loci for porcine skin thickness, we first performed a genome scan with 194 microsatellite markers in a White Duroc × Erhualian F₂ intercross. We identified three genome‐wide significant QTL on pig chromosomes (SSC) 4, 7 and 15 using linkage analysis. The most significant QTL was found on SSC7 with a small confidence interval of ~5 cM, explaining 23.9 percent of phenotypic variance. Further, we conducted a genome‐wide association study (GWAS) using Illumina PorcineSNP60 Beadchips for the F₂ pedigree and a population of Chinese Sutai pigs. We confirmed significant QTL in the F₂ pedigree and replicated QTL on SSC15 in Chinese Sutai pigs. A meta‐analysis of GWASs on both populations detected a genomic region associated with skin thickness on SSC4. GWAS results were generally consistent with QTL mapping. Identical‐by‐descent analysis defined QTL on SSC7 in a 683‐kb region harboring an interesting candidate gene: HMGA1. On SSC15, the linkage disequilibrium analysis showed a haplotype block of 2.20 Mb that likely harbors the gene responsible for skin thickness. Our findings provide novel insights into the genetic basis of swine skin thickness, which would benefit further understanding of porcine skin function.     

A genome‐wide association study reveals a quantitative trait locus for days open on chromosome 2 in Japanese Black cattle

Days open (DO), which is the interval from calving to conception, is an important trait related to reproductive performance in cattle. To identify quantitative trait loci for DO in Japanese Black cattle, we conducted a genome‐wide association study with 33 303 single nucleotide polymorphisms (SNPs) using 459 animals with extreme DO values selected from a larger group of 15 488 animals. We identified a SNP on bovine chromosome 2 (BTA2) that was associated with DO. After imputation using phased haplotype data inferred from 586 812 SNPs of 1041 Japanese Black cattle, six SNPs associated with DO were located in an 8.5‐kb region of high linkage disequilibrium on BTA2. These SNPs were located on the telomeric side at a distance of 177 kb from the parathyroid hormone 2 receptor (PTH2R) gene. The association was replicated in a sample of 1778 animals. In the replicated population, the frequency of the reduced‐DO allele (Q) was 0.63, and it accounted for 1.72% of the total genetic variance. The effect of a Q‐to‐q allele substitution on DO was a decrease of 3.74 days. The results suggest that the Q allele could serve as a marker in Japanese Black cattle to select animals with superior DO performance.     

A genome‐wide scan for copy number variations using high‐density single nucleotide polymorphism array in Simmental cattle

Copy number variations (CNVs) have recently been identified as promising sources of genetic variation, complementary to single nucleotide polymorphisms (SNPs). As a result, detection of CNVs has attracted a great deal of attention. In this study, we performed genome‐wide CNV detection using Illumina Bovine HD BeadChip (770k) data on 792 Simmental cattle. A total of 263 CNV regions (CNVRs) were identified, which included 137 losses, 102 gains and 24 regions classified as both loss and gain, covering 35.48 Mb (1.41%) of the bovine genome. The length of these CNVRs ranged from 10.18 kb to 1.76 Mb, with an average length of 134.78 kb and a median length of 61.95 kb. In 136 of these regions, a total of 313 genes were identified related to biological functions such as transmembrane activity and olfactory transduction activity. To validate the results, we performed quantitative PCR to detect nine randomly selected CNVRs and successfully confirmed seven (77.6%) of them. Our results present a map of cattle CNVs derived from high‐density SNP data, which expands the current CNV map of the cattle genome and provides useful information for investigation of genomic structural variation in cattle.     

Reducing the interval of a growth QTL on chromosome 4 in laying hens

In our previous research, we identified a QTL with an interval of 3.4 Mb for growth on chicken chromosome (GGA) 4 in an advanced intercross population of an initial cross between the New Hampshire inbred line (NHI) and the White Leghorn inbred line (WL77). In the current study, an association analysis was performed in a population of purebred white layers (WLA) with White Leghorn origin. Genotypic data of 130 SNPs within the previously identified 3.4‐Mb region were obtained using a 60K SNP chip. In total, 24 significant SNPs (LOD ≥ 4.44) on GGA4 were detected for daily weigh gain from 8 to 14 weeks and two SNPs (LOD ≥ 4.80) for body weight at 14 weeks. The QTL interval was reduced by 1.9 Mb to an interval of 1.5 Mb (74.6–76.1 Mb) that harbors 15 genes. Furthermore, to identify additional loci for chicken growth, a genome‐wide association study (GWAS) was carried out in a WLA population. The GWAS identified an additional QTL on GGA6 for body weight at six weeks (19.8–21.2 Mb). Our findings showed that by using a WLA population we were able to further reduce the QTL confidence interval previously detected using a NHI × WL77 advanced intercross population.     

Linked‐read sequencing enables haplotype‐resolved resequencing at population scale

The feasibility to sequence entire genomes of virtually any organism provides unprecedented insights into the evolutionary history of populations and species. Nevertheless, many population genomic inferences – including the quantification and dating of admixture, introgression and demographic events, and inference of selective sweeps – are still limited by the lack of high‐quality haplotype information. The newest generation of sequencing technology now promises significant progress. To establish the feasibility of haplotype‐resolved genome resequencing at population scale, we investigated properties of linked‐read sequencing data of songbirds of the genus Oenanthe across a range of sequencing depths. Our results based on the comparison of downsampled (25×, 20×, 15×, 10×, 7×, and 5×) with high‐coverage data (46–68×) of seven bird genomes mapped to a reference suggest that phasing contiguities and accuracies adequate for most population genomic analyses can be reached already with moderate sequencing effort. At 15× coverage, phased haplotypes span about 90% of the genome assembly, with 50% and 90% of phased sequences located in phase blocks longer than 1.25–4.6 Mb (N50) and 0.27–0.72 Mb (N90). Phasing accuracy reaches beyond 99% starting from 15× coverage. Higher coverages yielded higher contiguities (up to about 7 Mb/1 Mb [N50/N90] at 25× coverage), but only marginally improved phasing accuracy. Phase block contiguity improved with input DNA molecule length; thus, higher‐quality DNA may help keeping sequencing costs at bay. In conclusion, even for organisms with gigabase‐sized genomes like birds, linked‐read sequencing at moderate depth opens an affordable avenue towards haplotype‐resolved genome resequencing at population scale.     

Mapping QTL for production traits in segregating Piétrain pig populations using genome‐wide association study results of F2 crosses

In this study, genome‐wide association study (GWAS) results of porcine F2 crosses were used to map QTL in outcross Piétrain populations. For this purpose, two F2 crosses (Piétrain × Meishan, n = 304; Piétrain × Wild Boar, n = 291) were genotyped with the PorcineSNP60v2 BeadChip and phenotyped for the dressing yield, carcass length, daily gain and drip loss traits. GWASs were conducted in the pooled F2 cross applying single marker mixed linear models. For the investigated traits, between two and five (in total 15) QTL core regions, spanning 250 segregating SNPs around a significant trait‐associated peak SNP, were identified. The SNPs within the QTL core regions were subsequently tested for trait association in two outcross Piétrain populations consisting of 771 progeny‐tested boars and 210 sows with their own performance records. In the sow (boar) dataset, five (eight) of the 15 mapped QTL were validated. Hence, many QTL mapped in the F2 crosses (with Piétrain as a common founder breed) are still segregating in the current Piétrain breed. This confirms the usefulness of existing F2 crosses for mapping QTL that are still segregating in the recent founder breed generation. The approach utilizes the high power of an F2 cross to map QTL in a breeding population for which it is not guaranteed that they would be found using a GWAS in this population.     

Fine‐mapping the POLL locus in Brahman cattle yields the diagnostic marker CSAFG29

The POLL locus has been mapped to the centromeric region of bovine chromosome 1 (BTA1) in both taurine breeds and taurine–indicine crosses in an interval of approximately 1 Mb. It has not yet been mapped in pure‐bred zebu cattle. Despite several efforts, neither causative mutations in candidate genes nor a singular diagnostic DNA marker has been identified. In this study, we genotyped a total of 68 Brahman cattle and 20 Hereford cattle informative for the POLL locus for 33 DNA microsatellites, 16 of which we identified de novo from the bovine genome sequence, mapping the POLL locus to the region of the genes IFNAR2 and SYNJ1. The 303‐bp allele of the new microsatellite, CSAFG29, showed strong association with the POLL allele. We then genotyped 855 Brahman cattle for CSAFG29 and confirmed the association between the 303‐bp allele and POLL. To determine whether the same association was found in taurine breeds, we genotyped 334 animals of the Angus, Hereford and Limousin breeds and 376 animals of the Brangus, Droughtmaster and Santa Gertrudis composite taurine–zebu breeds. The association between the 303‐bp allele and POLL was confirmed in these breeds; however, an additional allele (305 bp) was also associated but not fully predictive of POLL. Across the data, CSAFG29 was in sufficient linkage disequilibrium to the POLL allele in Australian Brahman cattle that it could potentially be used as a diagnostic marker in that breed, but this may not be the case in other breeds. Further, we provide confirmatory evidence that the scur phenotype generally occurs in animals that are heterozygous for the POLL allele.     

Haplotype‐based genotyping‐by‐sequencing in oat genome research

In a de novo genotyping‐by‐sequencing (GBS) analysis of short, 64‐base tag‐level haplotypes in 4657 accessions of cultivated oat, we discovered 164741 tag‐level (TL) genetic variants containing 241224 SNPs. From this, the marker density of an oat consensus map was increased by the addition of more than 70000 loci. The mapped TL genotypes of a 635‐line diversity panel were used to infer chromosome‐level (CL) haplotype maps. These maps revealed differences in the number and size of haplotype blocks, as well as differences in haplotype diversity between chromosomes and subsets of the diversity panel. We then explored potential benefits of SNP vs. TL vs. CL GBS variants for mapping, high‐resolution genome analysis and genomic selection in oats. A combined genome‐wide association study (GWAS) of heading date from multiple locations using both TL haplotypes and individual SNP markers identified 184 significant associations. A comparative GWAS using TL haplotypes, CL haplotype blocks and their combinations demonstrated the superiority of using TL haplotype markers. Using a principal component‐based genome‐wide scan, genomic regions containing signatures of selection were identified. These regions may contain genes that are responsible for the local adaptation of oats to Northern American conditions. Genomic selection for heading date using TL haplotypes or SNP markers gave comparable and promising prediction accuracies of up to r = 0.74. Genomic selection carried out in an independent calibration and test population for heading date gave promising prediction accuracies that ranged between r = 0.42 and 0.67. In conclusion, TL haplotype GBS‐derived markers facilitate genome analysis and genomic selection in oat.     

Searching new signals for production traits through gene‐based association analysis in three Italian cattle breeds

Genome‐wide association studies (GWAS) have been widely applied to disentangle the genetic basis of complex traits. In cattle breeds, classical GWAS approaches with medium‐density marker panels are far from conclusive, especially for complex traits. This is due to the intrinsic limitations of GWAS and the assumptions that are made to step from the association signals to the functional variations. Here, we applied a gene‐based strategy to prioritize genotype–phenotype associations found for milk production and quality traits with classical approaches in three Italian dairy cattle breeds with different sample sizes (Italian Brown n = 745; Italian Holstein n = 2058; Italian Simmental n = 477). Although classical regression on single markers revealed only a single genome‐wide significant genotype–phenotype association, for Italian Holstein, the gene‐based approach identified specific genes in each breed that are associated with milk physiology and mammary gland development. As no standard method has yet been established to step from variation to functional units (i.e., genes), the strategy proposed here may contribute to revealing new genes that play significant roles in complex traits, such as those investigated here, amplifying low association signals using a gene‐centric approach.     

A genome‐wide association study suggests several novel candidate genes for carcass traits in Chinese Simmental beef cattle

A genome‐wide association study (GWAS) was conducted for two carcass traits in Chinese Simmental beef cattle. The experimental population consisted of 1301 individuals genotyped with the Illumina BovineHD SNP BeadChip (770K). After quality control, 671 990 SNPs and 1217 individuals were retained for the GWAS. The phenotypic traits included carcass weight and bone weight, which were measured after the cattle were slaughtered at 16 to 18 months of age. Three statistical models—a fixed polygene model, a random polygene model and a composite interval mapping polygene model—were used for the GWAS. The genome‐wide significance threshold after Bonferroni correction was 7.44E‐08 (= 0.05/671 990). In this study, we detected eight and seven SNPs significantly associated with carcass weight and bone weight respectively. In total, 11 candidate genes were identified within or close to these significant SNPs. Of these, we found several novel candidate genes, including PBX1, GCNT4, ALDH1A2, LCORL and WDFY3, to be associated with carcass weight and bone weight in Chinese Simmental beef cattle, and their functional roles need to be verified in further studies.     

Population genomics reveals a fine‐scale recombination landscape for genetic improvement of cotton

Recombination breaks up ancestral linkage disequilibrium, creates combinations of alleles, affects the efficiency of natural selection, and plays a major role in crop domestication and improvement. However, there is little knowledge regarding the variation in the population‐scaled recombination rate in cotton. We constructed recombination maps and characterized the difference in the genomic landscape of the population‐scaled recombination rate between Gossypium hirsutum and G. arboreum and sub‐genomes based on the 381 sequenced G. hirsutum and 215 G. arboreum accessions. Comparative genomics identified large structural variations and syntenic genes in the recombination regions, suggesting that recombination was related to structural variation and occurred preferentially in the distal chromosomal regions. Correlation analysis indicated that recombination was only slightly affected by geographical distribution and breeding period. A genome‐wide association study (GWAS) was performed with 15 agronomic traits using 267 cotton accessions and identified 163 quantitative trait loci (QTL) and an important candidate gene (Ghir_COL2) for early maturity traits. Comparative analysis of recombination and a GWAS revealed that the QTL of fibre quality traits tended to be more common in high‐recombination regions than were those of yield and early maturity traits. These results provide insights into the population‐scaled recombination landscape, suggesting that recombination contributed to the domestication and improvement of cotton, which provides a useful reference for studying recombination in other species.     

The pomegranate (Punica granatum L.) draft genome dissects genetic divergence between soft‐ and hard‐seeded cultivars

Complete and highly accurate reference genomes and gene annotations are indispensable for basic biological research and trait improvement of woody tree species. In this study, we integrated single‐molecule sequencing and high‐throughput chromosome conformation capture techniques to produce a high‐quality and long‐range contiguity chromosome‐scale genome assembly of the soft‐seeded pomegranate cultivar ‘Tunisia’. The genome covers 320.31 Mb (scaffold N50 = 39.96 Mb; contig N50 = 4.49 Mb) and includes 33 594 protein‐coding genes. We also resequenced 26 pomegranate varieties that varied regarding seed hardness. Comparative genomic analyses revealed many genetic differences between soft‐ and hard‐seeded pomegranate varieties. A set of selective loci containing SUC8‐like, SUC6, FoxO and MAPK were identified by the selective sweep analysis between hard‐ and soft‐seeded populations. An exceptionally large selective region (26.2 Mb) was identified on chromosome 1. Our assembled pomegranate genome is more complete than other currently available genome assemblies. Our results indicate that genomic variations and selective genes may have contributed to the genetic divergence between soft‐ and hard‐seeded pomegranate varieties.     

Large-effect pleiotropic or closely linked QTL segregate within and across ten US cattle breeds

Background

The availability of high-density SNP assays including the BovineSNP50 (50 K) enables the identification of novel quantitative trait loci (QTL) and improvement of the resolution of the locations of previously mapped QTL. We performed a series of genome-wide association studies (GWAS) using 50 K genotypes scored in 18,274 animals from 10 US beef cattle breeds with observations for twelve body weights, calving ease and carcass traits.

Results

A total of 159 large-effects QTL (defined as 1-Mb genome windows explaining more than 1% of additive genetic variance) were identified. In general, more QTL were identified in analyses with bigger sample sizes. Four large-effect pleiotropic or closely linked QTLs located on BTA6 at 37–42 Mb (primarily at 38 Mb), on BTA7 at 93 Mb, on BTA14 at 23–26 Mb (primarily at 25 Mb) and on BTA20 at 4 Mb were identified in more than one breed. Several breed-specific large-effect pleiotropic or closely linked QTL were also identified. Some identified QTL regions harbor genes known to have large effects on a variety of traits in cattle such as PLAG1 and MSTN and others harbor promising candidate genes including NCAPG, ARRDC3, ERGIC1, SH3PXD2B, HMGA2, MSRB3, LEMD3, TIGAR, SEPT7, and KIRREL3. Gene ontology analysis revealed that genes involved in ossification and in adipose tissue development were over-represented in the identified pleiotropic QTL. Also, the MAPK signaling pathway was identified as a common pathway affected by the genes located near the pleiotropic QTL.

Conclusions

This largest GWAS ever performed in beef cattle, led us to discover several novel across-breed and breed-specific large-effect pleiotropic QTL that cumulatively account for a significant percentage of additive genetic variance (e.g. more than a third of additive genetic variance of birth and mature weights; and calving ease direct in Hereford). These results will improve our understanding of the biology of growth and body composition in cattle.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2164-15-442) contains supplementary material, which is available to authorized users.     

Grazing behaviour of dairy cows on biodiverse mountain pastures is more influenced by slope than cow breed

The aim of this study was to determine how cows with different genetic merit behave and perform when grazing biodiverse and heterogeneous mountain pastures with different slopes. Three groups of 12 cows in late lactation, each composed of four Holstein, four Montbéliarde and four Valdostana Red Pied cows, breeds of increasing presumed robustness and decreasing milk yield (MY) potential. Cows grazed without concentrate either on a low-diversity flat pasture or on two species-rich mountainous pastures having slopes of either 7° or 22°. Milk yield, BW and grazing behaviour were monitored two times in the first and once in the second grazing cycle. Cows of different breeds had similar behaviour on all pastures. The Montbéliarde cows performed close to their production potential; Holstein and Valdostana cows produced less milk than anticipated. No breed difference in terms of BW loss was found. The Valdostana cows exhibited the least selective behaviour with respect to plant species and plant growth stage. Still, all cows searched for the most palatable vegetation regardless of pasture diversity. On the steep pasture, cows optimised the trade-off between ingesting and saving energy to obtain feed. They remained longer at the lowest zone and selected forbs, whereas cows on the flatter pasture went to the upper zone to select grasses. The present study gave no evidence for a superior short-term adaptation to harsh grazing conditions through an optimised feeding behaviour of the Valdostana breed compared to Montbéliarde and Holstein cows.     

Genome‐wide association analyses reveal the genetic basis of combining ability in rice

Combining ability is a measure for selecting elite parents and predicting hybrid performance in plant breeding. However, the genetic basis of combining ability remains unclear and a global view of combining ability from diverse mating designs is lacking. We developed a North Carolina II (NCII) population of 96 Oryza sativa and four male sterile lines to identify parents of greatest value for hybrid rice production. Statistical analyses indicated that general combining ability (GCA) and specific combining ability (SCA) contributed variously to different agronomic traits. In a genome‐wide association study (GWAS) of agronomic traits, GCA and SCA, we identified 34 significant associations (P < 2.39 × 10^?7). The superior alleles of GCA loci (Ghd8, GS3 and qSSR4) accumulated in parental lines with high GCA and explained 30.03% of GCA variance in grain yield, indicating that molecular breeding of high GCA parental lines is feasible. The distinct distributions of these QTLs contributed to the differentiation of parental GCA in subpopulations. GWAS of SCA identified 12 more loci that showed dominance on corresponding agronomic traits. We conclude that the accumulation of superior GCA and SCA alleles is an important contributor to heterosis and QTLs that greatly contributed to combining ability in our study would accelerate the identification of elite inbred lines and breeding of super hybrids.