Using genomic prediction with crop growth models enables the prediction of associated traits in wheat.

Crop growth models (CGM) can predict the performance of a cultivar in untested environments by sampling genotype-specific parameters. As they cannot predict the performance of new cultivars, it has been proposed to integrate CGMs with whole genome prediction (WGP) to combine the benefits of both models. Here, we used a CGM-WGP model to predict the performance of new wheat (Triticum aestivum) genotypes. The CGM was designed to predict phenology, nitrogen, and biomass traits. The CGM-WGP model simulated more heritable GSPs compared with the CGM and gave smaller errors for the observed phenotypes. The WGP model performed better when predicting yield, grain number, and grain protein content, but showed comparable performance to the CGM-WGP model for heading and physiological maturity dates. However, the CGM-WGP model was able to predict unobserved traits (for which there were no phenotypic records in the reference population). The CGM-WGP model also showed superior performance when predicting unrelated individuals that clustered separately from the reference population. Our results demonstrate new advantages for CGM-WGP modelling and suggest future efforts should focus on calibrating CGM-WGP models using high-throughput phenotypic measures that are cheaper and less laborious to collect.

Integrating biophysical crop growth models and whole genome prediction for their mutual benefit: a case study in wheat phenology.

Running crop growth models (CGM) coupled with whole genome prediction (WGP) as a CGM-WGP model introduces environmental information to WGP and genomic relatedness information to the genotype-specific parameters modelled through CGMs. Previous studies have primarily used CGM-WGP to infer prediction accuracy without exploring its potential to enhance CGM and WGP. Here, we implemented a heading and maturity date wheat phenology model within a CGM-WGP framework and compared it with CGM and WGP. The CGM-WGP resulted in more heritable genotype-specific parameters with more biologically realistic correlation structures between genotype-specific parameters and phenology traits compared with CGM-modelled genotype-specific parameters that reflected the correlation of measured phenotypes. Another advantage of CGM-WGP is the ability to infer accurate prediction with much smaller and less diverse reference data compared with that required for CGM. A genome-wide association analysis linked the genotype-specific parameters from the CGM-WGP model to nine significant phenology loci including Vrn-A1 and the three PPD1 genes, which were not detected for CGM-modelled genotype-specific parameters. Selection on genotype-specific parameters could be simpler than on observed phenotypes. For example, thermal time traits are theoretically more independent candidates, compared with the highly correlated heading and maturity dates, which could be used to achieve an environment-specific optimal flowering period. CGM-WGP combines the advantages of CGM and WGP to predict more accurate phenotypes for new genotypes under alternative or future environmental conditions.

In it for the long run: perspectives on exploiting long-read sequencing in livestock for population scale studies of structural variants.

Studies have demonstrated that structural variants (SV) play a substantial role in the evolution of species and have an impact on Mendelian traits in the genome. However, unlike small variants (< 50 bp), it has been challenging to accurately identify and genotype SV at the population scale using short-read sequencing. Long-read sequencing technologies are becoming competitively priced and can address several of the disadvantages of short-read sequencing for the discovery and genotyping of SV. In livestock species, analysis of SV at the population scale still faces challenges due to the lack of resources, high costs, technological barriers, and computational limitations. In this review, we summarize recent progress in the characterization of SV in the major livestock species, the obstacles that still need to be overcome, as well as the future directions in this growing field. It seems timely that research communities pool resources to build global population-scale long-read sequencing consortiums for the major livestock species for which the application of genomic tools has become cost-effective.

Demographic History, Adaptation, and NRAP Convergent Evolution at Amino Acid Residue 100 in the World Northernmost Cattle from Siberia.

Native cattle breeds represent an important cultural heritage. They are a reservoir of genetic variation useful for properly responding to agriculture needs in the light of ongoing climate changes. Evolutionary processes that occur in response to extreme environmental conditions could also be better understood using adapted local populations. Herein, different evolutionary histories of the world northernmost native cattle breeds from Russia were investigated. They highlighted Kholmogory as a typical taurine cattle, whereas Yakut cattle separated from European taurines approximately 5,000 years ago and contain numerous ancestral and some novel genetic variants allowing their adaptation to harsh conditions of living above the Polar Circle. Scans for selection signatures pointed to several common gene pathways related to adaptation to harsh climates in both breeds. But genes affected by selection from these pathways were mostly different. A Yakut cattle breed-specific missense mutation in a highly conserved NRAP gene represents a unique example of a young amino acid residue convergent change shared with at least 16 species of hibernating/cold-adapted mammals from six distinct phylogenetic orders. This suggests a convergent evolution event along the mammalian phylogenetic tree and fast fixation in a single isolated cattle population exposed to a harsh climate.

Strategies of preserving genetic diversity while maximizing genetic response from implementing genomic selection in pulse breeding programs.

KEY MESSAGE: Genomic selection maximizes genetic gain by recycling parents to germplasm pool earlier and preserves genetic diversity by restricting the number of fixed alleles and the relationship in pulse breeding programs. Using a stochastic computer simulation, we investigated the benefit of optimization strategies in the context of genomic selection (GS) for pulse breeding programs. We simulated GS for moderately complex to highly complex traits such as disease resistance, grain weight and grain yield in multiple environments with a high level of genotype-by-environment interaction for grain yield. GS led to higher genetic gain per unit of time and higher genetic diversity loss than phenotypic selection by shortening the breeding cycle time. The genetic gain obtained from selecting the segregating parents early in the breeding cycle (at F1 or F2 stages) was substantially higher than selecting at later stages even though prediction accuracy was moderate. Increasing the number of F1 intercross (F1i) families and keeping the total number of progeny of F1i families constant, we observed a decrease in genetic gain and increase in genetic diversity, whereas increasing the number of progeny per F1i family while keeping a constant number of F1i families increased the rate of genetic gain and had higher genetic diversity loss per unit of time. Adding 50 F2 family phenotypes to the training population increased the accuracy of genomic breeding values (GEBVs) and genetic gain per year and decreased the rate of genetic diversity loss. Genetic diversity could be preserved by applying a strategy that restricted both the percentage of alleles fixed and the average relationship of the group of selected parents to preserve long-term genetic improvement in the pulse breeding program.

Genome-wide association studies dissect the G × E interaction for agronomic traits in a worldwide collection of safflowers (Carthamus tinctorius L.).

Genome-wide association studies were conducted using a globally diverse safflower (Carthamus tinctorius L.) Genebank collection for grain yield (YP), days to flowering (DF), plant height (PH), 500 seed weight (SW), seed oil content (OL), and crude protein content (PR) in four environments (sites) that differed in water availability. Phenotypic variation was observed for all traits. YP exhibited low overall genetic correlations (rGoverall) across sites, while SW and OL had high rGoverall and high pairwise genetic correlations (rGij) across all pairwise sites. In total, 92 marker-trait associations (MTAs) were identified using three methods, single locus genome-wide association studies (GWAS) using a mixed linear model (MLM), the Bayesian multi-locus method (BayesR), and meta-GWAS. MTAs with large effects across all sites were detected for OL, SW, and PR, and MTAs specific for the different water stress sites were identified for all traits. Five MTAs were associated with multiple traits; 4 of 5 MTAs were variously associated with the three traits of SW, OL, and PR. This study provided insights into the phenotypic variability and genetic architecture of important safflower agronomic traits under different environments. Supplementary Information: The online version contains supplementary material available at 10.1007/s11032-022-01295-8.

Quantifying the contribution of sequence variants with regulatory and evolutionary significance to 34 bovine complex traits.

Many genome variants shaping mammalian phenotype are hypothesized to regulate gene transcription and/or to be under selection. However, most of the evidence to support this hypothesis comes from human studies. Systematic evidence for regulatory and evolutionary signals contributing to complex traits in a different mammalian model is needed. Sequence variants associated with gene expression (expression quantitative trait loci [eQTLs]) and concentration of metabolites (metabolic quantitative trait loci [mQTLs]) and under histone-modification marks in several tissues were discovered from multiomics data of over 400 cattle. Variants under selection and evolutionary constraint were identified using genome databases of multiple species. These analyses defined 30 sets of variants, and for each set, we estimated the genetic variance the set explained across 34 complex traits in 11,923 bulls and 32,347 cows with 17,669,372 imputed variants. The per-variant trait heritability of these sets across traits was highly consistent (r > 0.94) between bulls and cows. Based on the per-variant heritability, conserved sites across 100 vertebrate species and mQTLs ranked the highest, followed by eQTLs, young variants, those under histone-modification marks, and selection signatures. From these results, we defined a Functional-And-Evolutionary Trait Heritability (FAETH) score indicating the functionality and predicted heritability of each variant. In additional 7,551 cattle, the high FAETH-ranking variants had significantly increased genetic variances and genomic prediction accuracies in 3 production traits compared to the low FAETH-ranking variants. The FAETH framework combines the information of gene regulation, evolution, and trait heritability to rank variants, and the publicly available FAETH data provide a set of biological priors for cattle genomic selection worldwide.

Meta-analysis of genome-wide association studies reveal common loci controlling agronomic and quality traits in a wide range of normal and heat stressed environments.

KEY MESSAGE: Several stable QTL were detected using metaGWAS analysis for different agronomic and quality traits under 26 normal and heat stressed environments. Heat stress, exacerbated by global warming, has a negative influence on wheat production worldwide and climate resilient cultivars can help mitigate these impacts. Selection decisions should therefore depend on multi-environment experiments representing a range of temperatures at critical stages of development. Here, we applied a meta-genome wide association analysis (metaGWAS) approach to detect stable QTL with significant effects across multiple environments. The metaGWAS was applied to 11 traits scored in 26 trials that were sown at optimal or late times of sowing (TOS1 and TOS2, respectively) at five locations. A total of 2571 unique wheat genotypes (13,959 genotypes across all environments) were included and the analysis conducted on TOS1, TOS2 and both times of sowing combined (TOS1&2). The germplasm was genotyped using a 90 k Infinium chip and imputed to exome sequence level, resulting in 341,195 single nucleotide polymorphisms (SNPs). The average accuracy across all imputed SNPs was high (92.4%). The three metaGWAS analyses revealed 107 QTL for the 11 traits, of which 16 were detected in all three analyses and 23 were detected in TOS1&2 only. The remaining QTL were detected in either TOS1 or TOS2 with or without TOS1&2, reflecting the complex interactions between the environments and the detected QTL. Eight QTL were associated with grain yield and seven with multiple traits. The identified QTL provide an important resource for gene enrichment and fine mapping to further understand the mechanisms of gene × environment interaction under both heat stressed and unstressed conditions.

Genomic selection can accelerate the biofortification of spring wheat.

KEY MESSAGE: Genomic selection enabled accurate prediction for the concentration of 13 nutritional element traits in wheat. Wheat biofortification is one of the most sustainable strategies to alleviate mineral deficiency in human diets. Here, we investigated the potential of genomic selection using BayesR and Bayesian ridge regression (BRR) models to predict grain yield (YLD) and the concentration of 13 nutritional elements in grains (B, Ca, Co, Cu, Fe, K, Mg, Mn, Mo, Na, Ni, P and Zn) using a population of 1470 spring wheat lines. The lines were grown in replicated field trials with two times of sowing (TOS) at 3 locations (Narrabri-NSW, all lines; Merredin-WA and Horsham-VIC, 200 core lines). Narrow-sense heritability across environments (locations/TOS) ranged from 0.09 to 0.45. Co, K, Na and Ca showed low to negative genetic correlations with other traits including YLD, while the remaining traits were negatively correlated with YLD. When all environments were included in the reference population, medium to high prediction accuracy was observed for the different traits across environments. BayesR had higher average prediction accuracy for mineral concentrations (r = 0.55) compared to BRR (r = 0.48) across all traits and environments but both methods had comparable accuracies for YLD. We also investigated the utility of one or two locations (reference locations) to predict the remaining location(s), as well as the ability of one TOS to predict the other. Under these scenarios, BayesR and BRR showed comparable performance but with lower prediction accuracy compared to the scenario of predicting reference environments for new lines. Our study demonstrates the potential of genomic selection for enriching wheat grain with nutritional elements in biofortification breeding.

A conditional multi-trait sequence GWAS discovers pleiotropic candidate genes and variants for sheep wool, skin wrinkle and breech cover traits.

BACKGROUND: Imputation to whole-genome sequence is now possible in large sheep populations. It is therefore of interest to use this data in genome-wide association studies (GWAS) to investigate putative causal variants and genes that underpin economically important traits. Merino wool is globally sought after for luxury fabrics, but some key wool quality attributes are unfavourably correlated with the characteristic skin wrinkle of Merinos. In turn, skin wrinkle is strongly linked to susceptibility to "fly strike" (Cutaneous myiasis), which is a major welfare issue. Here, we use whole-genome sequence data in a multi-trait GWAS to identify pleiotropic putative causal variants and genes associated with changes in key wool traits and skin wrinkle. RESULTS: A stepwise conditional multi-trait GWAS (CM-GWAS) identified putative causal variants and related genes from 178 independent quantitative trait loci (QTL) of 16 wool and skin wrinkle traits, measured on up to 7218 Merino sheep with 31 million imputed whole-genome sequence (WGS) genotypes. Novel candidate gene findings included the MAT1A gene that encodes an enzyme involved in the sulphur metabolism pathway critical to production of wool proteins, and the ESRP1 gene. We also discovered a significant wrinkle variant upstream of the HAS2 gene, which in dogs is associated with the exaggerated skin folds in the Shar-Pei breed. CONCLUSIONS: The wool and skin wrinkle traits studied here appear to be highly polygenic with many putative candidate variants showing considerable pleiotropy. Our CM-GWAS identified many highly plausible candidate genes for wool traits as well as breech wrinkle and breech area wool cover.

Expression quantitative trait loci in sheep liver and muscle contribute to variations in meat traits.

BACKGROUND: Variants that regulate transcription, such as expression quantitative trait loci (eQTL), have shown enrichment in genome-wide association studies (GWAS) for mammalian complex traits. However, no study has reported eQTL in sheep, although it is an important agricultural species for which many GWAS of complex meat traits have been conducted. Using RNA sequence data produced from liver and muscle from 149 sheep and imputed whole-genome single nucleotide polymorphisms (SNPs), our aim was to dissect the genetic architecture of the transcriptome by associating sheep genotypes with three major molecular phenotypes including gene expression (geQTL), exon expression (eeQTL) and RNA splicing (sQTL). We also examined these three types of eQTL for their enrichment in GWAS of multi-meat traits and fatty acid profiles. RESULTS: Whereas a relatively small number of molecular phenotypes were significantly heritable (h2 > 0, P < 0.05), their mean heritability ranged from 0.67 to 0.73 for liver and from 0.71 to 0.77 for muscle. Association analysis between molecular phenotypes and SNPs within ± 1 Mb identified many significant cis-eQTL (false discovery rate, FDR < 0.01). The median distance between the eQTL and transcription start sites (TSS) ranged from 68 to 153 kb across the three eQTL types. The number of common variants between geQTL, eeQTL and sQTL within each tissue, and the number of common variants between liver and muscle within each eQTL type were all significantly (P < 0.05) larger than expected by chance. The identified eQTL were significantly (P < 0.05) enriched in GWAS hits associated with 56 carcass traits and fatty acid profiles. For example, several geQTL in muscle mapped to the FAM184B gene, hundreds of sQTL in liver and muscle mapped to the CAST gene, and hundreds of sQTL in liver mapped to the C6 gene. These three genes are associated with body composition or fatty acid profiles. CONCLUSIONS: We detected a large number of significant eQTL and found that the overlap of variants between eQTL types and tissues was prevalent. Many eQTL were also QTL for meat traits. Our study fills a gap in the knowledge on the regulatory variants and their role in complex traits for the sheep model.

Mitochondrial protein gene expression and the oxidative phosphorylation pathway associated with feed efficiency and energy balance in dairy cattle.

Feed efficiency and energy balance are important traits underpinning profitability and environmental sustainability in animal production. They are complex traits, and our understanding of their underlying biology is currently limited. One measure of feed efficiency is residual feed intake (RFI), which is the difference between actual and predicted intake. Variation in RFI among individuals is attributable to the metabolic efficiency of energy utilization. High RFI (H_RFI) animals require more energy per unit of weight gain or milk produced compared with low RFI (L_RFI) animals. Energy balance (EB) is a closely related trait calculated very similarly to RFI. Cellular energy metabolism in mitochondria involves mitochondrial protein (MiP) encoded by both nuclear (NuMiP) and mitochondrial (MtMiP) genomes. We hypothesized that MiP genes are differentially expressed (DE) between H_RFI and L_RFI animal groups and similarly between negative and positive EB groups. Our study aimed to characterize MiP gene expression in white blood cells of H_RFI and L_RFI cows using RNA sequencing to identify genes and biological pathways associated with feed efficiency in dairy cattle. We used the top and bottom 14 cows ranked for RFI and EB out of 109 animals as H_RFI and L_RFI, and positive and negative EB groups, respectively. The gene expression counts across all nuclear and mitochondrial genes for animals in each group were used for differential gene expression analyses, weighted gene correlation network analysis, functional enrichment, and identification of hub genes. Out of 244 DE genes between RFI groups, 38 were MiP genes. The DE genes were enriched for the oxidative phosphorylation (OXPHOS) and ribosome pathways. The DE MiP genes were underexpressed in L_RFI (and negative EB) compared with the H_RFI (and positive EB) groups, suggestive of reduced mitochondrial activity in the L_RFI group. None of the MtMiP genes were among the DE MiP genes between the groups, which suggests a non-rate limiting role of MtMiP genes in feed efficiency and warrants further investigation. The role of MiP, particularly the NuMiP and OXPHOS pathways in RFI, was also supported by our gene correlation network analysis and the hub gene identification. We validated the findings in an independent data set. Overall, our study suggested that differences in feed efficiency in dairy cows may be linked to differences in cellular energy demand. This study broadens our knowledge of the biology of feed efficiency in dairy cattle.

Population-dependent reproducible deviation from natural bread wheat genome in synthetic hexaploid wheat.

Sequence elimination is one of the main mechanisms that increases the divergence among homoeologous chromosomes after allopolyploidization to enhance the stability of recently established lineages, but it can cause a loss of some economically important genes. Synthetic hexaploid wheat (SHW) is an important source of genetic variation to the natural hexaploid wheat (NHW) genepool that has low genetic diversity. Here, we investigated the change between SHW and NHW genomes by utilizing a large germplasm set of primary synthetics and synthetic derivatives. Reproducible segment elimination (RSE) was declared if a large chromosomal chunk (>5 cM) produced no aligned reads in more than five SHWs. RSE in five genomic regions was the major source of variation between SHW and NHW. One RSE eliminated almost the complete short arm of chromosome 1B, which contains major genes for flour quality, disease resistance and different enzymes. The occurrence of RSE was highly dependent on the choice of diploid and tetraploid parental lines, their ancestral subpopulation and admixture, e.g. SHWs derived from Triticum dicoccon or from one of two Aegilops tauschii subpopulations were almost free of RSE, while highly admixed parents had higher RSE rates. The rate of RSE in synthetic derivatives was almost double that in primary synthetics. Genome-wide association analysis detected four loci with minor effects on the occurrence of RSE, indicating that both parental lines and genetic factors were affecting the occurrence of RSE. Therefore, pre-pre-breeding strategies should be applied before introducing SHW into pre-breeding programs to ensure genomic stability and avoid undesirable gene loss.

Expression of mitochondrial protein genes encoded by nuclear and mitochondrial genomes correlate with energy metabolism in dairy cattle.

BACKGROUND: Mutations in the mitochondrial genome have been implicated in mitochondrial disease, often characterized by impaired cellular energy metabolism. Cellular energy metabolism in mitochondria involves mitochondrial proteins (MP) from both the nuclear (NuMP) and mitochondrial (MtMP) genomes. The expression of MP genes in tissues may be tissue specific to meet varying specific energy demands across the tissues. Currently, the characteristics of MP gene expression in tissues of dairy cattle are not well understood. In this study, we profile the expression of MP genes in 29 adult and six foetal tissues in dairy cattle using RNA sequencing and gene expression analyses: particularly differential gene expression and co-expression network analyses. RESULTS: MP genes were differentially expressed (DE; over-expressed or under-expressed) across tissues in cattle. All 29 tissues showed DE NuMP genes in varying proportions of over-expression and under-expression. On the other hand, DE of MtMP genes was observed in < 50% of tissues and notably MtMP genes within a tissue was either all over-expressed or all under-expressed. A high proportion of NuMP (up to 60%) and MtMP (up to 100%) genes were over-expressed in tissues with expected high metabolic demand; heart, skeletal muscles and tongue, and under-expressed (up to 45% of NuMP, 77% of MtMP genes) in tissues with expected low metabolic rates; leukocytes, thymus, and lymph nodes. These tissues also invariably had the expression of all MtMP genes in the direction of dominant NuMP genes expression. The NuMP and MtMP genes were highly co-expressed across tissues and co-expression of genes in a cluster were non-random and functionally enriched for energy generation pathway. The differential gene expression and co-expression patterns were validated in independent cow and sheep datasets. CONCLUSIONS: The results of this study support the concept that there are biological interaction of MP genes from the mitochondrial and nuclear genomes given their over-expression in tissues with high energy demand and co-expression in tissues. This highlights the importance of considering MP genes from both genomes in future studies related to mitochondrial functions and traits related to energy metabolism.

Selection signatures in tropical cattle are enriched for promoter and coding regions and reveal missense mutations in the damage response gene HELB.

BACKGROUND: Distinct domestication events, adaptation to different climatic zones, and divergent selection in productive traits have shaped the genomic differences between taurine and indicine cattle. In this study, we assessed the impact of artificial selection and environmental adaptation by comparing whole-genome sequences from European taurine and Asian indicine breeds and from African cattle. Next, we studied the impact of divergent selection by exploiting predicted and experimental functional annotation of the bovine genome. RESULTS: We identified selective sweeps in beef cattle taurine and indicine populations, including a 430-kb selective sweep on indicine cattle chromosome 5 that is located between 47,670,001 and 48,100,000 bp and spans five genes, i.e. HELB, IRAK3, ENSBTAG00000026993, GRIP1 and part of HMGA2. Regions under selection in indicine cattle display significant enrichment for promoters and coding genes. At the nucleotide level, sites that show a strong divergence in allele frequency between European taurine and Asian indicine are enriched for the same functional categories. We identified nine single nucleotide polymorphisms (SNPs) in coding regions that are fixed for different alleles between subspecies, eight of which were located within the DNA helicase B (HELB) gene. By mining information from the 1000 Bull Genomes Project, we found that HELB carries mutations that are specific to indicine cattle but also found in taurine cattle, which are known to have been subject to indicine introgression from breeds, such as N'Dama, Anatolian Red, Marchigiana, Chianina, and Piedmontese. Based on in-house genome sequences, we proved that mutations in HELB segregate independently of the copy number variation HMGA2-CNV, which is located in the same region. CONCLUSIONS: Major genomic sequence differences between Bos taurus and Bos indicus are enriched for promoter and coding regions. We identified a 430-kb selective sweep in Asian indicine cattle located on chromosome 5, which carries SNPs that are fixed in indicine populations and located in the coding sequences of the HELB gene. HELB is involved in the response to DNA damage including exposure to ultra-violet light and is associated with reproductive traits and yearling weight in tropical cattle. Thus, HELB likely contributed to the adaptation of tropical cattle to their harsh environment.

Analyses of inter-individual variations of sperm DNA methylation and their potential implications in cattle.

BACKGROUND: DNA methylation has been shown to be involved in many biological processes, including X chromosome inactivation in females, paternal genomic imprinting, and others. RESULTS: Based on the correlation patterns of methylation levels of neighboring CpG sites among 28 sperm whole genome bisulfite sequencing (WGBS) data (486 × coverage), we obtained 31,272 methylation haplotype blocks (MHBs). Among them, we defined conserved methylated regions (CMRs), variably methylated regions (VMRs) and highly variably methylated regions (HVMRs) among individuals, and showed that HVMRs might play roles in transcriptional regulation and function in complex traits variation and adaptive evolution by integrating evidence from traditional and molecular quantitative trait loci (QTL), and selection signatures. Using a weighted correlation network analysis (WGCNA), we also detected a co-regulated module of HVMRs that was significantly associated with reproduction traits, and enriched for glycosyltransferase genes, which play critical roles in spermatogenesis and fertilization. Additionally, we identified 46 VMRs significantly associated with reproduction traits, nine of which were regulated by cis-SNPs, implying the possible intrinsic relationships among genomic variations, DNA methylation, and phenotypes. These significant VMRs were co-localized (± 10 kb) with genes related to sperm motility and reproduction, including ZFP36L1, CRISP2 and HGF. We provided further evidence that rs109326022 within a predominant QTL on BTA18 might influence the reproduction traits through regulating the methylation level of nearby genes JOSD2 and ASPDH in sperm. CONCLUSION: In summary, our results demonstrated associations of sperm DNA methylation with reproduction traits, highlighting the potential of epigenomic information in genomic improvement programs for cattle.

Extension of a haplotype-based genomic prediction model to manage multi-environment wheat data using environmental covariates.

KEY MESSAGE: A multi-environment genomic prediction model incorporating environmental covariates increased the prediction accuracy of wheat grain protein content. The advantage of the haplotype-based model was dependent upon the trait of interest. The inclusion of environment covariates (EC) in genomic prediction models has the potential to precisely model environmental effects and genotype-by-environment interactions. Together with EC, a haplotype-based genomic prediction approach, which is capable of accommodating the interaction between local epistasis and environment, may increase the prediction accuracy. The main objectives of our study were to evaluate the potential of EC to portray the relationship between environments and the relevance of local epistasis modelled by haplotype-based approaches in multi-environment prediction. The results showed that among five traits: grain yield (GY), plant height, protein content, screenings percentage (SP) and thousand kernel weight, protein content exhibited a 2.1% increase in prediction accuracy when EC was used to model the environmental relationship compared to treatment of the environment as a regular random effect without a variance-covariance structure. The approach used a Gaussian kernel to characterise the relationship among environments that displayed no advantage in contrast to the use of a genomic relationship matrix. The prediction accuracies of haplotype-based approaches for SP were consistently higher than the genotype-based model when the numbers of single-nucleotide polymorphisms (SNP) in a haplotype were from three to ten. In contrast, for GY, haplotype-based models outperformed genotype-based methods when two to four SNPs were used to construct the haplotype.

Genomic prediction based on selected variants from imputed whole-genome sequence data in Australian sheep populations.

BACKGROUND: Whole-genome sequence (WGS) data could contain information on genetic variants at or in high linkage disequilibrium with causative mutations that underlie the genetic variation of polygenic traits. Thus far, genomic prediction accuracy has shown limited increase when using such information in dairy cattle studies, in which one or few breeds with limited diversity predominate. The objective of our study was to evaluate the accuracy of genomic prediction in a multi-breed Australian sheep population of relatively less related target individuals, when using information on imputed WGS genotypes. METHODS: Between 9626 and 26,657 animals with phenotypes were available for nine economically important sheep production traits and all had WGS imputed genotypes. About 30% of the data were used to discover predictive single nucleotide polymorphism (SNPs) based on a genome-wide association study (GWAS) and the remaining data were used for training and validation of genomic prediction. Prediction accuracy using selected variants from imputed sequence data was compared to that using a standard array of 50k SNP genotypes, thereby comparing genomic best linear prediction (GBLUP) and Bayesian methods (BayesR/BayesRC). Accuracy of genomic prediction was evaluated in two independent populations that were each lowly related to the training set, one being purebred Merino and the other crossbred Border Leicester x Merino sheep. RESULTS: A substantial improvement in prediction accuracy was observed when selected sequence variants were fitted alongside 50k genotypes as a separate variance component in GBLUP (2GBLUP) or in Bayesian analysis as a separate category of SNPs (BayesRC). From an average accuracy of 0.27 in both validation sets for the 50k array, the average absolute increase in accuracy across traits with 2GBLUP was 0.083 and 0.073 for purebred and crossbred animals, respectively, whereas with BayesRC it was 0.102 and 0.087. The average gain in accuracy was smaller when selected sequence variants were treated in the same category as 50k SNPs. Very little improvement over 50k prediction was observed when using all WGS variants. CONCLUSIONS: Accuracy of genomic prediction in diverse sheep populations increased substantially by using variants selected from whole-genome sequence data based on an independent multi-breed GWAS, when compared to genomic prediction using standard 50K genotypes.

Using imputed whole-genome sequence data to improve the accuracy of genomic prediction for parasite resistance in Australian sheep.

BACKGROUND: This study aimed at (1) comparing the accuracies of genomic prediction for parasite resistance in sheep based on whole-genome sequence (WGS) data to those based on 50k and high-density (HD) single nucleotide polymorphism (SNP) panels; (2) investigating whether the use of variants within quantitative trait loci (QTL) regions that were selected from regional heritability mapping (RHM) in an independent dataset improved the accuracy more than variants selected from genome-wide association studies (GWAS); and (3) comparing the prediction accuracies between variants selected from WGS data to variants selected from the HD SNP panel. RESULTS: The accuracy of genomic prediction improved marginally from 0.16 ± 0.02 and 0.18 ± 0.01 when using all the variants from 50k and HD genotypes, respectively, to 0.19 ± 0.01 when using all the variants from WGS data. Fitting a GRM from the selected variants alongside a GRM from the 50k SNP genotypes improved the prediction accuracy substantially compared to fitting the 50k SNP genotypes alone. The gain in prediction accuracy was slightly more pronounced when variants were selected from WGS data compared to when variants were selected from the HD panel. When sequence variants that passed the GWAS [Formula: see text] threshold of 3 across the entire genome were selected, the prediction accuracy improved by 5% (up to 0.21 ± 0.01), whereas when selection was limited to sequence variants that passed the same GWAS [Formula: see text] threshold of 3 in regions identified by RHM, the accuracy improved by 9% (up to 0.25 ± 0.01). CONCLUSIONS: Our results show that through careful selection of sequence variants from the QTL regions, the accuracy of genomic prediction for parasite resistance in sheep can be improved. These findings have important implications for genomic prediction in sheep.

Accuracy of imputation to whole-genome sequence in sheep.

BACKGROUND: The use of whole-genome sequence (WGS) data for genomic prediction and association studies is highly desirable because the causal mutations should be present in the data. The sequencing of 935 sheep from a range of breeds provides the opportunity to impute sheep genotyped with single nucleotide polymorphism (SNP) arrays to WGS. This study evaluated the accuracy of imputation from SNP genotypes to WGS using this reference population of 935 sequenced sheep. RESULTS: The accuracy of imputation from the Ovine Infinium® HD BeadChip SNP (~ 500 k) to WGS was assessed for three target breeds: Merino, Poll Dorset and F1 Border Leicester × Merino. Imputation accuracy was highest for the Poll Dorset breed, although there were more Merino individuals in the sequenced reference population than Poll Dorset individuals. In addition, empirical imputation accuracies were higher (by up to 1.7%) when using larger multi-breed reference populations compared to using a smaller single-breed reference population. The mean accuracy of imputation across target breeds using the Minimac3 or the FImpute software was 0.94. The empirical imputation accuracy varied considerably across the genome; six chromosomes carried regions of one or more Mb with a mean imputation accuracy of < 0.7. Imputation accuracy in five variant annotation classes ranged from 0.87 (missense) up to 0.94 (intronic variants), where lower accuracy corresponded to higher proportions of rare alleles. The imputation quality statistic reported from Minimac3 (R2) had a clear positive relationship with the empirical imputation accuracy. Therefore, by first discarding imputed variants with an R2 below 0.4, the mean empirical accuracy across target breeds increased to 0.97. Although accuracy of genomic prediction was less affected by filtering on R2 in a multi-breed population of sheep with imputed WGS, the genomic heritability clearly tended to be lower when using variants with an R2 ≤ 0.4. CONCLUSIONS: The mean imputation accuracy was high for all target breeds and was increased by combining smaller breed sets into a multi-breed reference. We found that the Minimac3 software imputation quality statistic (R2) was a useful indicator of empirical imputation accuracy, enabling removal of very poorly imputed variants before downstream analyses.