Increased genomic prediction accuracy in wheat breeding through spatial adjustment of field trial data

Genomic selection in plant breeding: Key factors shaping two decades of progress

Genomic selection, the application of genomic prediction (GP) models to select candidate individuals, has significantly advanced in the past two decades, effectively accelerating genetic gains in plant breeding. This article provides a holistic overview of key factors that have influenced GP in plant breeding during this period. We delved into the pivotal roles of training population size and genetic diversity, and their relationship with the breeding population, in determining GP accuracy. Special emphasis was placed on optimizing training population size. We explored its benefits and the associated diminishing returns beyond an optimum size. This was done while considering the balance between resource allocation and maximizing prediction accuracy through current optimization algorithms. The density and distribution of single-nucleotide polymorphisms, level of linkage disequilibrium, genetic complexity, trait heritability, statistical machine-learning methods, and non-additive effects are the other vital factors. Using wheat, maize, and potato as examples, we summarize the effect of these factors on the accuracy of GP for various traits. The search for high accuracy in GP-theoretically reaching one when using the Pearson's correlation as a metric-is an active research area as yet far from optimal for various traits. We hypothesize that with ultra-high sizes of genotypic and phenotypic datasets, effective training population optimization methods and support from other omics approaches (transcriptomics, metabolomics and proteomics) coupled with deep-learning algorithms could overcome the boundaries of current limitations to achieve the highest possible prediction accuracy, making genomic selection an effective tool in plant breeding.

Selective Genotyping and Phenotyping for Optimization of Genomic Prediction Models for Populations with Different Diversity

To overcome the different challenges to food security caused by a growing population and climate change, soybean (Glycine max (L.) Merr.) breeders are creating novel cultivars that have the potential to improve productivity while maintaining environmental sustainability. Genomic selection (GS) is an advanced approach that may accelerate the rate of genetic gain in breeding using genome-wide molecular markers. The accuracy of genomic selection can be affected by trait architecture and heritability, marker density, linkage disequilibrium, statistical models, and training set. The selection of a minimal and optimal marker set with high prediction accuracy can lower genotyping costs, computational time, and multicollinearity. Selective phenotyping could reduce the number of genotypes tested in the field while preserving the genetic diversity of the initial population. This study aimed to evaluate different methods of selective genotyping and phenotyping on the accuracy of genomic prediction for soybean yield. The evaluation was performed on three populations: recombinant inbred lines, multifamily diverse lines, and germplasm collection. Strategies adopted for marker selection were as follows: SNP (single nucleotide polymorphism) pruning, estimation of marker effects, randomly selected markers, and genome-wide association study. Reduction of the number of genotypes was performed by selecting a core set from the initial population based on marker data, yet maintaining the original population's genetic diversity. Prediction ability using all markers and genotypes was different among examined populations. The subsets obtained by the model-based strategy can be considered the most suitable for marker selection for all populations. The selective phenotyping based on makers in all cases had higher values of prediction ability compared to minimal values of prediction ability of multiple cycles of random selection, with the highest values of prediction obtained using AN approach and 75% population size. The obtained results indicate that selective genotyping and phenotyping hold great potential and can be integrated as tools for improving or retaining selection accuracy by reducing genotyping or phenotyping costs for genomic selection.

Genome-wide association study reveals GmFulb as candidate gene for maturity time and reproductive length in soybeans (Glycine max)

The ability of soybean [Glycine max (L.) Merr.] to adapt to different latitudes is attributed to genetic variation in major E genes and quantitative trait loci (QTLs) determining flowering time (R1), maturity (R8), and reproductive length (RL). Fully revealing the genetic basis of R1, R8, and RL in soybeans is necessary to enhance genetic gains in soybean yield improvement. Here, we performed a genome-wide association analysis (GWA) with 31,689 single nucleotide polymorphisms (SNPs) to detect novel loci for R1, R8, and RL using a soybean panel of 329 accessions with the same genotype for three major E genes (e1-as/E2/E3). The studied accessions were grown in nine environments and observed for R1, R8 and RL in all environments. This study identified two stable peaks on Chr 4, simultaneously controlling R8 and RL. In addition, we identified a third peak on Chr 10 controlling R1. Association peaks overlap with previously reported QTLs for R1, R8, and RL. Considering the alternative alleles, significant SNPs caused RL to be two days shorter, R1 two days later and R8 two days earlier, respectively. We identified association peaks acting independently over R1 and R8, suggesting that trait-specific minor effect loci are also involved in controlling R1 and R8. From the 111 genes highly associated with the three peaks detected in this study, we selected six candidate genes as the most likely cause of R1, R8, and RL variation. High correspondence was observed between a modifying variant SNP at position 04:39294836 in GmFulb and an association peak on Chr 4. Further studies using map-based cloning and fine mapping are necessary to elucidate the role of the candidates we identified for soybean maturity and adaptation to different latitudes and to be effectively used in the marker-assisted breeding of cultivars with optimal yield-related traits.

Haplotype blocks for genomic prediction: a comparative evaluation in multiple crop datasets

In modern plant breeding, genomic selection is becoming the gold standard for selection of superior genotypes. The basis for genomic prediction models is a set of phenotyped lines along with their genotypic profile. With high marker density and linkage disequilibrium (LD) between markers, genotype data in breeding populations tends to exhibit considerable redundancy. Therefore, interest is growing in the use of haplotype blocks to overcome redundancy by summarizing co-inherited features. Moreover, haplotype blocks can help to capture local epistasis caused by interacting loci. Here, we compared genomic prediction methods that either used single SNPs or haplotype blocks with regards to their prediction accuracy for important traits in crop datasets. We used four published datasets from canola, maize, wheat and soybean. Different approaches to construct haplotype blocks were compared, including blocks based on LD, physical distance, number of adjacent markers and the algorithms implemented in the software "Haploview" and "HaploBlocker". The tested prediction methods included Genomic Best Linear Unbiased Prediction (GBLUP), Extended GBLUP to account for additive by additive epistasis (EGBLUP), Bayesian LASSO and Reproducing Kernel Hilbert Space (RKHS) regression. We found improved prediction accuracy in some traits when using haplotype blocks compared to SNP-based predictions, however the magnitude of improvement was very trait- and model-specific. Especially in settings with low marker density, haplotype blocks can improve genomic prediction accuracy. In most cases, physically large haplotype blocks yielded a strong decrease in prediction accuracy. Especially when prediction accuracy varies greatly across different prediction models, prediction based on haplotype blocks can improve prediction accuracy of underperforming models. However, there is no "best" method to build haplotype blocks, since prediction accuracy varied considerably across methods and traits. Hence, criteria used to define haplotype blocks should not be viewed as fixed biological parameters, but rather as hyperparameters that need to be adjusted for every dataset.

Digitalizing breeding in plants: A new trend of next-generation breeding based on genomic prediction

As the world's population grows and food needs diversification, the demand for cereals and horticultural crops with beneficial traits increases. In order to meet a variety of demands, suitable cultivars and innovative breeding methods need to be developed. Breeding methods have changed over time following the advance of genetics. With the advent of new sequencing technology in the early 21st century, predictive breeding, such as genomic selection (GS), emerged when large-scale genomic information became available. GS shows good predictive ability for the selection of individuals with traits of interest even for quantitative traits by using various types of the whole genome-scanning markers, breaking away from the limitations of marker-assisted selection (MAS). In the current review, we briefly describe the history of breeding techniques, each breeding method, various statistical models applied to GS and methods to increase the GS efficiency. Consequently, we intend to propose and define the term digital breeding through this review article. Digital breeding is to develop a predictive breeding methods such as GS at a higher level, aiming to minimize human intervention by automatically proceeding breeding design, propagating breeding populations, and to make selections in consideration of various environments, climates, and topography during the breeding process. We also classified the phases of digital breeding based on the technologies and methods applied to each phase. This review paper will provide an understanding and a direction for the final evolution of plant breeding in the future.

Spectral-Based Classification of Genetically Differentiated Groups in Spring Wheat Grown under Contrasting Environments

The global concern about the gap between food production and consumption has intensified the research on the genetics, ecophysiology, and breeding of cereal crops. In this sense, several genetic studies have been conducted to assess the effectiveness and sustainability of collections of germplasm accessions of major crops. In this study, a spectral-based classification approach for the assignment of wheat cultivars to genetically differentiated subpopulations (genetic structure) was carried out using a panel of 316 spring bread cultivars grown in two environments with different water regimes (rainfed and fully irrigated). For that, different machine-learning models were trained with foliar spectral and genetic information to assign the wheat cultivars to subpopulations. The results revealed that, in general, the hyperparameters ReLU (as the activation function), adam (as the optimizer), and a size batch of 10 give neural network models better accuracy. Genetically differentiated groups showed smaller differences in mean wavelengths under rainfed than under full irrigation, which coincided with a reduction in clustering accuracy in neural network models. The comparison of models indicated that the Convolutional Neural Network (CNN) was significantly more accurate in classifying individuals into their respective subpopulations, with 92 and 93% of correct individual assignments in water-limited and fully irrigated environments, respectively, whereas 92% (full irrigation) and 78% (rainfed) of cultivars were correctly assigned to their respective classes by the multilayer perceptron method and partial least squares discriminant analysis, respectively. Notably, CNN did not show significant differences between both environments, which indicates stability in the prediction independent of the different water regimes. It is concluded that foliar spectral variation can be used to accurately infer the belonging of a cultivar to its respective genetically differentiated group, even considering radically different environments, which is highly desirable in the context of crop genetic resources management.

Physiological breeding for yield improvement in soybean: solar radiation interception-conversion, and harvest index

KEY MESSAGE

Efficiency of light interception, Radiation use efficiency and harvest index can be used as targets to improve grain yield potential in soybean. Grain yield (GY) production can be expressed as the result of three main efficiencies: light interception (Ei), radiation use (RUE), and harvest index (HI). Although dissecting GY through these three efficiencies is not entirely new, there is a lack of knowledge about the phenotypic variation, the genetic architecture, and the relative contribution of these three efficiencies on GY in soybean. This knowledge gap coupled with laborious phenotyping prevents the active consideration of these efficiencies into breeding programs. This study aims to reveal the phenotypic variation, heritability, genetic relationships, genetic architecture, and genomic prediction for Ei, RUE, and HI in soybean. We evaluated a maturity control panel of 383 Recombinant Inbred Lines (RILs) selected from the soybean nested association mapping (SoyNAM) population. Dry matter ground measured along with canopy coverage (CC) from UAS imagery were collected in three environments. Light interception was modeled through a logistic curve using CC as a proxy. The total above-ground biomass collected during the growing season and its respective cumulative light intercepted were used to derive RUE through linear models fitting. Additive-genetic correlations, genome-wide association (GWA) and whole-genome regressions (WGR) were performed to evaluate the relationship between traits, their association with genomic regions, and the feasibility of predicting these efficiencies with genomic information. Correlation analyses considered three groups: the entire data set, and the high- and low-yielding RILs to determine association as a function of the GY. Our results revealed moderate to high phenotypic variation for Ei, RUE, and HI with ranges of 8.5%, 1.1 g MJ-1, and 0.2, respectively. Additive-genetic correlation revealed a strong relationship of GY with HI and moderate with RUE and Ei when whole data set was considered, but negligible contribution of HI on GY when just the top 100 was analyzed. The GWA analyses showed that Ei is associated with three SNPs; two of them located on chromosome 7 and one on chromosome 11 with no previous quantitative trait loci (QTLs) reported for these regions. RUE is associated with four SNPs on chromosomes 1, 7, 11, and 18. Some of these QTLs are novel, while others are previously documented for plant architecture and chlorophyll content. Two SNPs positioned on chromosome 13 and 15 with previous QTLs reported for plant height and seed set, weight and abortion were associated with HI. WGR showed high predictive ability for Ei, RUE, and HI with maximum correlation ranging between 0.75 and 0.80. Future improvements in GY can be expected through strategies prioritizing Ei for short-term results when using high yielding germplasm and RUE for medium- and long-term outcomes. This work is a pioneer attempt to integrate traditional physiological traits into the breeding process in the context of physiological breeding.

Spatial Regression Models for Field Trials: A Comparative Study and New Ideas

Naturally occurring variability within a study region harbors valuable information on relationships between biological variables. Yet, spatial patterns within these study areas, e.g., in field trials, violate the assumption of independence of observations, setting particular challenges in terms of hypothesis testing, parameter estimation, feature selection, and model evaluation. We evaluate a number of spatial regression methods in a simulation study, including more realistic spatial effects than employed so far. Based on our results, we recommend generalized least squares (GLS) estimation for experimental as well as for observational setups and demonstrate how it can be incorporated into popular regression models for high-dimensional data such as regularized least squares. This new method is available in the BioConductor R-package pengls. Inclusion of a spatial error structure improves parameter estimation and predictive model performance in low-dimensional settings and also improves feature selection in high-dimensional settings by reducing "red-shift": the preferential selection of features with spatial structure. In addition, we argue that the absence of spatial autocorrelation (SAC) in the model residuals should not be taken as a sign of a good fit, since it may result from overfitting the spatial trend. Finally, we confirm our findings in a case study on the prediction of winter wheat yield based on multispectral measurements.

Multi-Trait Multi-Environment Genomic Prediction for End-Use Quality Traits in Winter Wheat

Soft white wheat is a wheat class used in foreign and domestic markets to make various end products requiring specific quality attributes. Due to associated cost, time, and amount of seed needed, phenotyping for the end-use quality trait is delayed until later generations. Previously, we explored the potential of using genomic selection (GS) for selecting superior genotypes earlier in the breeding program. Breeders typically measure multiple traits across various locations, and it opens up the avenue for exploring multi-trait-based GS models. This study's main objective was to explore the potential of using multi-trait GS models for predicting seven different end-use quality traits using cross-validation, independent prediction, and across-location predictions in a wheat breeding program. The population used consisted of 666 soft white wheat genotypes planted for 5 years at two locations in Washington, United States. We optimized and compared the performances of four uni-trait- and multi-trait-based GS models, namely, Bayes B, genomic best linear unbiased prediction (GBLUP), multilayer perceptron (MLP), and random forests. The prediction accuracies for multi-trait GS models were 5.5 and 7.9% superior to uni-trait models for the within-environment and across-location predictions. Multi-trait machine and deep learning models performed superior to GBLUP and Bayes B for across-location predictions, but their advantages diminished when the genotype by environment component was included in the model. The highest improvement in prediction accuracy, that is, 35% was obtained for flour protein content with the multi-trait MLP model. This study showed the potential of using multi-trait-based GS models to enhance prediction accuracy by using information from previously phenotyped traits. It would assist in speeding up the breeding cycle time in a cost-friendly manner.

Insights Into the Genetic Architecture of Complex Traits in Napier Grass (Cenchrus purpureus) and QTL Regions Governing Forage Biomass Yield, Water Use Efficiency and Feed Quality Traits

Napier grass is the most important perennial tropical grass native to Sub-Saharan Africa and widely grown in tropical and subtropical regions around the world, primarily as a forage crop for animal feed, but with potential as an energy crop and in a wide range of other areas. Genomic resources have recently been developed for Napier grass that need to be deployed for genetic improvement and molecular dissection of important agro-morphological and feed quality traits. From a diverse set of Napier grass genotypes assembled from two independent collections, a subset of 84 genotypes (although a small population size, the genotypes were selected to best represent the genetic diversity of the collections) were selected and evaluated for 2 years in dry (DS) and wet (WS) seasons under three soil moisture conditions: moderate water stress in DS (DS-MWS); severe water stress in DS (DS-SWS) and, under rainfed (RF) conditions in WS (WS-RF). Data for agro-morphological and feed quality traits, adjusted for the spatial heterogeneity in the experimental blocks, were collected over a 2-year period from 2018 to 2020. A total of 135,706 molecular markers were filtered, after removing markers with missing values >10% and a minor allele frequency (MAF) <5%, from the high-density genome-wide markers generated previously using the genotyping by sequencing (GBS) method of the DArTseq platform. A genome-wide association study (GWAS), using two different mixed linear model algorithms implemented in the GAPIT R package, identified more than 35 QTL regions and markers associated with agronomic, morphological, and water-use efficiency traits. QTL regions governing purple pigmentation and feed quality traits were also identified. The identified markers will be useful in the genetic improvement of Napier grass through the application of marker-assisted selection and for further characterization and map-based cloning of the QTLs.

Accounting for spatial trends in multi-environment diallel analysis in maize breeding

Spatial trends represent an obstacle to genetic evaluation in maize breeding. Spatial analyses can correct spatial trends, which allow for an increase in selective accuracy. The objective of this study was to compare the spatial (SPA) and non-spatial (NSPA) models in diallel multi-environment trial analyses in maize breeding. The trials consisted of 78 inter-populational maize hybrids, tested in four environments (E1, E2, E3, and E4), with three replications, under a randomized complete block design. The SPA models accounted for autocorrelation among rows and columns by the inclusion of first-order autoregressive matrices (AR1 ⊗ AR1). Then, the rows and columns factors were included in the fixed and random parts of the model. Based on the Bayesian information criteria, the SPA models were used to analyze trials E3 and E4, while the NSPA model was used for analyzing trials E1 and E2. In the joint analysis, the compound symmetry structure for the genotypic effects presented the best fit. The likelihood ratio test showed that some effects changed regarding significance when the SPA and NSPA models were used. In addition, the heritability, selective accuracy, and selection gain were higher when the SPA models were used. This indicates the power of the SPA model in dealing with spatial trends. The SPA model exhibits higher reliability values and is recommended to be incorporated in the standard procedure of genetic evaluation in maize breeding. The analyses bring the parents 2, 10 and 12, as potential parents in this microregion.

Integrating genomic selection with a genotype plus genotype x environment (GGE) model improves prediction accuracy and computational efficiency

Genotype-by-environment interaction (GEI) is one of the major factors affecting the prediction accuracy of genomic selection (GS) models. Standard models have low power to model complex GEI, and they fail to predict phenotypes in unobserved environments. Here, we developed a novel prediction model that account for GEI, named 3GS, that combines genotype plus genotype × environment (GGE) analysis with GS. The model calculates the principal components (PCs) of the environmental phenotypes using GGE analysis and predict the performance of these PCs using standard GS models before converting the GEBVs of these PCs (pcGEBVs) back to the original phenotypes. We demonstrated three advantages of the new model. First, 3GS showed significantly higher prediction accuracy primarily for deviated environments that have low to negative correlations to other environments. Second, 3GS can predict new genotypes in unobserved environments with high accuracy. Third, the computational complexity of 3GS increases linearly with increasing the number of environments and the population size, unlike the standard models that exhibit exponential increase, making it hundreds of times faster than the standard models for large data sets. 3GS can improve prediction accuracy with minimal resources in modern breeding programmes in which massive populations get multi-environment phenotypes with high-throughput techniques.

Spatial kernel models capturing field heterogeneity for accurate estimation of genetic potential

According to Fisher's principles, an experimental field is typically divided into multiple blocks for local control. Although homogeneity is supposed within a block, this assumption may not be practical for large blocks, such as those including hundreds of plots. In line evaluation trials, which are essential in plant breeding, field heterogeneity must be carefully treated, because it can cause bias in the estimation of genetic potential. To more accurately estimate genotypic values in a large field trial, we developed spatial kernel models incorporating genome-wide markers, which consider continuous heterogeneity within a block and over the field. In the simulation study, the spatial kernel models were robust under various conditions. Although heritability, spatial autocorrelation range, replication number, and missing plots directly affected the estimation accuracy of genotypic values, the spatial kernel models always showed superior performance over the classical block model. We also employed these spatial kernel models for quantitative trait locus mapping. Finally, using field experimental data of bioenergy sorghum lines, we validated the performance of the spatial kernel models. The results suggested that a spatial kernel model is effective for evaluating the genetic potential of lines in a heterogeneous field.

Development of whole-genome prediction models to increase the rate of genetic gain in intermediate wheatgrass (Thinopyrum intermedium) breeding

The development of perennial grain crops is driven by the vision of simultaneous food production and enhanced ecosystem services. Typically, perennial crops like intermediate wheatgrass (IWG)[Thinopyrum intermedium (Host) Barkworth & D.R Dewey] have low seed yield and other detrimental traits. Next-generation sequencing has made genomic selection (GS) a tractable and viable breeding method. To investigate how an IWG breeding program may use GS, we evaluated 3,658 genets over 2 yr for 46 traits to build a training population. Six statistical models were used to evaluate the non-replicated data, and a model using autoregressive order 1 (AR1) spatial correction for rows and columns combined with the genomic relationship matrix provided the highest estimates of heritability. Genomic selection models were built from 18,357 single nucleotide polymorphism markers via genotyping-by-sequencing, and a 20-fold cross-validation showed high predictive ability for all traits (r > .80). Predictive abilities improved with increased training population size and marker numbers, even with larger amounts of missing data per marker. On the basis of these results, we propose a GS breeding method that is capable of completing one cycle per year compared with a minimum of 2 yr per cycle with phenotypic selection. We estimate that this breeding approach can increase the rate of genetic gain up to 2.6× above phenotypic selection for spike yield in IWG, allowing GS to enable rapid domestication and improvement of this crop. These breeding methods should be transferable to other species with similar long breeding cycles or limited capacity for replicated observations.

Genomics and breeding innovations for enhancing genetic gain for climate resilience and nutrition traits

KEY MESSAGE

Integrating genomics technologies and breeding methods to tweak core parameters of the breeder's equation could accelerate delivery of climate-resilient and nutrient rich crops for future food security. Accelerating genetic gain in crop improvement programs with respect to climate resilience and nutrition traits, and the realization of the improved gain in farmers' fields require integration of several approaches. This article focuses on innovative approaches to address core components of the breeder's equation. A prerequisite to enhancing genetic variance (σ2g) is the identification or creation of favorable alleles/haplotypes and their deployment for improving key traits. Novel alleles for new and existing target traits need to be accessed and added to the breeding population while maintaining genetic diversity. Selection intensity (i) in the breeding program can be improved by testing a larger population size, enabled by the statistical designs with minimal replications and high-throughput phenotyping. Selection priorities and criteria to select appropriate portion of the population too assume an important role. The most important component of breeder's equation is heritability (h2). Heritability estimates depend on several factors including the size and the type of population and the statistical methods. The present article starts with a brief discussion on the potential ways to enhance σ2g in the population. We highlight statistical methods and experimental designs that could improve trait heritability estimation. We also offer a perspective on reducing the breeding cycle time (t), which could be achieved through the selection of appropriate parents, optimizing the breeding scheme, rapid fixation of target alleles, and combining speed breeding with breeding programs to optimize trials for release. Finally, we summarize knowledge from multiple disciplines for enhancing genetic gains for climate resilience and nutritional traits.

Genetic Relationships Among Physiological Processes, Phenology, and Grain Yield Offer an Insight Into the Development of New Cultivars in Soybean (Glycine max L. Merr)

Soybean grain yield has steadily increased during the last century because of enhanced cultivars and better agronomic practices. Increases in the total biomass, shorter cultivars, late maturity, and extended seed-filling period are frequently reported as main contributors for better soybean performance. However, there are still processes associated with crop physiology to be improved. From the theoretical standpoint, yield is the product of efficiency of light interception (Ei), radiation use efficiency (RUE), and harvest index (HI). The relative contribution of these three parameters on the final grain yield (GY), their interrelation with other phenological-physiological traits, and their environmental stability have not been well established for soybean. In this study, we determined the additive-genetic relationship among 14 physiological and phenological traits including photosynthesis (A) and intrinsic water use efficiency (iWUE) in a panel of 383 soybean recombinant inbred lines (RILs) through direct (path analyses) and indirect learning methods [least absolute shrinkage and selection operator (LASSO) algorithm]. We evaluated the stability of Ei, RUE, and HI through the slope from the Finley and Wilkinson joint regression and the genetic correlation between traits evaluated in different environments. Results indicate that both supervised and unsupervised methods effectively establish the main relationships underlying changes in Ei, RUE, HI, and GY. Variations in the average growth rate of canopy coverage for the first 40 days after planting (AGR40) explain most of the changes in Ei. RUE is primarily influenced by phenological traits of reproductive length (RL) and seed-filling (SFL) as well as iWUE, light extinction coefficient (K), and A. HI showed a strong relationship with A, AGR40, SFL, and RL. According to the path analysis, an increase in one standard unit of HI promotes changes in 0.5 standard units of GY, while changes in the same standard unit of RUE and Ei produce increases on GY of 0.20 and 0.19 standard units, respectively. RUE, Ei, and HI exhibited better environmental stability than GY, although changes associated with year and location showed a moderate effect in Ei and RUE, respectively. This study brings insight into a group of traits involving A, iWUE, and RL to be prioritized during the breeding process for high-yielding cultivars.

Genome-Wide Association of Stem Carbohydrate Accumulation and Remobilization during Grain Growth in Bread Wheat (Triticum aestivum L.) in Mediterranean Environments

Water deficit represents an important challenge for wheat production in many regions of the world. Accumulation and remobilization of water-soluble carbohydrates (WSCs) in stems are part of the physiological responses regulated by plants to cope with water stress and, in turn, determine grain yield (GY). The genetic mechanisms underlying the variation in WSC are only partially understood. In this study, we aimed to identify Single Nucleotide Polymorphism (SNP) markers that account for variation in a suite of WSC and GY, evaluated in 225 cultivars and advanced lines of spring wheat. These genotypes were established in two sites in the Mediterranean region of Central Chile, under water-limited and full irrigation conditions, and assessed in two growing seasons, namely anthesis and maturity growth periods. A genome-wide association study (GWAS) was performed by using 3243 SNP markers. Genetic variance accounted for 5 to 52% of phenotypic variation of the assessed traits. A rapid linkage disequilibrium decay was observed across chromosomes (r² ≤ 0.2 at 2.52 kbp). Marker-trait association tests identified 96 SNPs related to stem weight (SW), WSCs, and GY, among other traits, at the different sites, growing seasons, and growth periods. The percentage of SNPs that were part of the gene-coding regions was 34%. Most of these genes are involved in the defensive response to drought and biotic stress. A complimentary analysis detected significant effects of different haplotypes on WSC and SW, in anthesis and maturity. Our results evidence both genetic and environmental influence on WSC dynamics in spring wheat. At the same time, they provide a series of markers suitable for supporting assisted selection approaches and functional characterization of genes.

Natural variation and genomic prediction of growth, physiological traits, and nitrogen-use efficiency in perennial ryegrass under low-nitrogen stress

Genomic prediction of nitrogen-use efficiency (NUE) has not previously been studied in perennial grass species exposed to low-N stress. Here, we conducted a genomic prediction of physiological traits and NUE in 184 global accessions of perennial ryegrass (Lolium perenne) in response to a normal (7.5 mM) and low (0.75 mM) supply of N. After 21 d of treatment under greenhouse conditions, significant variations in plant height increment (ΔHT), leaf fresh weight (LFW), leaf dry weight (LDW), chlorophyll index (Chl), chlorophyll fluorescence, leaf N and carbon (C) contents, C/N ratio, and NUE were observed in accessions , but to a greater extent under low-N stress. Six genomic prediction models were applied to the data, namely the Bayesian method Bayes C, Bayesian LASSO, Bayesian Ridge Regression, Ridge Regression-Best Linear Unbiased Prediction, Reproducing Kernel Hilbert Spaces, and randomForest. These models produced similar prediction accuracy of traits within the normal or low-N treatments, but the accuracy differed between the two treatments. ΔHT, LFW, LDW, and C were predicted slightly better under normal N with a mean Pearson r-value of 0.26, compared with r=0.22 under low N, while the prediction accuracies for Chl, N, C/N, and NUE were significantly improved under low-N stress with a mean r=0.45, compared with r=0.26 under normal N. The population panel contained three population structures, which generally had no effect on prediction accuracy. The moderate prediction accuracies obtained for N, C, and NUE under low-N stress are promising, and suggest a feasible means by which germplasm might be initially assessed for further detailed studies in breeding programs.

Quantitative Genomic Dissection of Soybean Yield Components

Soybean is a crop of major economic importance with low rates of genetic gains for grain yield compared to other field crops. A deeper understanding of the genetic architecture of yield components may enable better ways to tackle the breeding challenges. Key yield components include the total number of pods, nodes and the ratio pods per node. We evaluated the SoyNAM population, containing approximately 5600 lines from 40 biparental families that share a common parent, in 6 environments distributed across 3 years. The study indicates that the yield components under evaluation have low heritability, a reasonable amount of epistatic control, and partially oligogenic architecture: 18 quantitative trait loci were identified across the three yield components using multi-approach signal detection. Genetic correlation between yield and yield components was highly variable from family-to-family, ranging from -0.2 to 0.5. The genotype-by-environment correlation of yield components ranged from -0.1 to 0.4 within families. The number of pods can be utilized for indirect selection of yield. The selection of soybean for enhanced yield components can be successfully performed via genomic prediction, but the challenging data collections necessary to recalibrate models over time makes the introgression of QTL a potentially more feasible breeding strategy. The genomic prediction of yield components was relatively accurate across families, but less accurate predictions were obtained from within family predictions and predicting families not observed included in the calibration set.

Genomic prediction of yield and root development in wheat under changing water availability

BACKGROUND

Deeper roots help plants take up available resources in deep soil ensuring better growth and higher yields under conditions of drought. A large-scale semi-field root phenotyping facility was developed to allow a water availability gradient and detect potential interaction of genotype by water availability gradient. Genotyped winter wheat lines were grown as rows in four beds of this facility, where indirect genetic effects from neighbors could be important to trait variation. The objective was to explore the possibility of genomic prediction for grain-related traits and deep root traits collected via images taken in a minirhizotron tube under each row of winter wheat measured.

RESULTS

The analysis comprised four grain-related traits: grain yield, thousand-kernel weight, protein concentration, and total nitrogen content measured on each half row that were harvested separately. Two root traits, total root length between 1.2 and 2 m depth and root length in four intervals on each tube were also analyzed. Two sets of models with or without the effects of neighbors from both sides of each row were applied. No interaction between genotypes and changing water availability were detected for any trait. Estimated genomic heritabilities ranged from 0.263 to 0.680 for grain-related traits and from 0.030 to 0.055 for root traits. The coefficients of genetic variation were similar for grain-related and root traits. The prediction accuracy of breeding values ranged from 0.440 to 0.598 for grain-related traits and from 0.264 to 0.334 for root traits. Including neighbor effects in the model generally increased the estimated genomic heritabilities and accuracy of predicted breeding values for grain yield and nitrogen content.

CONCLUSIONS

Similar relative amounts of additive genetic variance were found for both yield traits and root traits but no interaction between genotypes and water availability were detected. It is possible to obtain accurate genomic prediction of breeding values for grain-related traits and reasonably accurate predicted breeding values for deep root traits using records from the semi-field facility. Including neighbor effects increased the estimated additive genetic variance of grain-related traits and accuracy of predicting breeding values. High prediction accuracy can be obtained although heritability is low.

Flexible modelling of spatial variation in agricultural field trials with the R package INLA

KEY MESSAGE

Established spatial models improve the analysis of agricultural field trials with or without genomic data and can be fitted with the open-source R package INLA. The objective of this paper was to fit different established spatial models for analysing agricultural field trials using the open-source R package INLA. Spatial variation is common in field trials, and accounting for it increases the accuracy of estimated genetic effects. However, this is still hindered by the lack of available software implementations. We compare some established spatial models and show possibilities for flexible modelling with respect to field trial design and joint modelling over multiple years and locations. We use a Bayesian framework and for statistical inference the integrated nested Laplace approximations (INLA) implemented in the R package INLA. The spatial models we use are the well-known independent row and column effects, separable first-order autoregressive ([Formula: see text]) models and a Gaussian random field (Matérn) model that is approximated via the stochastic partial differential equation approach. The Matérn model can accommodate flexible field trial designs and yields interpretable parameters. We test the models in a simulation study imitating a wheat breeding programme with different levels of spatial variation, with and without genome-wide markers and with combining data over two locations, modelling spatial and genetic effects jointly. The results show comparable predictive performance for both the [Formula: see text] and the Matérn models. We also present an example of fitting the models to a real wheat breeding data and simulated tree breeding data with the Nelder wheel design to show the flexibility of the Matérn model and the R package INLA.

Improving the efficiency of soybean breeding with high-throughput canopy phenotyping

BACKGROUND

In the early stages of plant breeding programs high-quality phenotypes are still a constraint to improve genetic gain. New field-based high-throughput phenotyping (HTP) platforms have the capacity to rapidly assess thousands of plots in a field with high spatial and temporal resolution, with the potential to measure secondary traits correlated to yield throughout the growing season. These secondary traits may be key to select more time and most efficiently soybean lines with high yield potential. Soybean average canopy coverage (ACC), measured by unmanned aerial systems (UAS), is highly heritable, with a high genetic correlation with yield. The objective of this study was to compare the direct selection for yield with indirect selection using ACC and using ACC as a covariate in the yield prediction model (Yield|ACC) in early stages of soybean breeding. In 2015 and 2016 we grew progeny rows (PR) and collected yield and days to maturity (R8) in a typical way and canopy coverage using a UAS carrying an RGB camera. The best soybean lines were then selected with three parameters, Yield, ACC and Yield|ACC, and advanced to preliminary yield trials (PYT).

RESULTS

We found that for the PYT in 2016, after adjusting yield for R8, there was no significant difference among the mean performances of the lines selected based on ACC and Yield. In the PYT in 2017 we found that the highest yield mean was from the lines directly selected for yield, but it may be due to environmental constraints in the canopy growth. Our results indicated that PR selection using Yield|ACC selected the most top-ranking lines in advanced yield trials.

CONCLUSIONS

Our findings emphasize the value of aerial HTP platforms for early stages of plant breeding. Though ACC selection did not result in the best performance lines in the second year of selections, our results indicate that ACC has a role in the effective selection of high-yielding soybean lines.

bWGR: Bayesian Whole-Genome Regression

MOTIVATION

Whole-genome regressions methods represent a key framework for genome-wide prediction, cross-validation studies, and association analysis. The bWGR offers a compendium of Bayesian methods with various priors available, allowing users to predict complex traits with different genetic architectures.

RESULTS

Here we introduce bWGR, an R package that enables users to efficient fit and cross-validate Bayesian and likelihood whole-genome regression methods. It implements a series of methods referred to as the Bayesian alphabet under the traditional Gibbs sampling and optimized Expectation-Maximization. The package also enables fitting efficient multivariate models and complex hierarchical models. The package is user-friendly and computational efficient.

AVAILABILITY AND IMPLEMENTATION

bWGR is an R package available in the CRAN repository. It can be installed in R by typing: install.packages("bWGR").

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

A metabolome-based core hybridisation strategy for the prediction of rice grain weight across environments

Marker-based prediction holds great promise for improving current plant and animal breeding efficiencies. However, the predictabilities of complex traits are always severely affected by negative factors, including distant relatedness, environmental discrepancies, unknown population structures, and indeterminate numbers of predictive variables. In this study, we utilised two independent F1 hybrid populations in the years 2012 and 2015 to predict rice thousand grain weight (TGW) using parental untargeted metabolite profiles with a partial least squares regression method. A stable predictive model for TGW was built based on hybrids from the population in 2012 (r = 0.75) but failed to properly predict TGW for hybrids from the population in 2015 (r = 0.27). After integrating hybrids from both populations into the training set, the TGW of hybrids could be predicted but was largely dependent on population structures. Then, core hybrids from each population were determined by principal component analysis and the TGW of hybrids in both environments were successfully predicted (r > 0.60). Moreover, adjusting the population structures and numbers of predictive analytes increased TGW predictability for hybrids in 2015 (r = 0.72). Our study demonstrates that the TGW of F1 hybrids across environments can be accurately predicted based on parental untargeted metabolite profiles with a core hybridisation strategy in rice. Metabolic biomarkers identified from early developmental stage tissues, which are grown under experimental conditions, may represent a workable approach towards the robust prediction of major agronomic traits for climate-adaptive varieties.

Design of training populations for selective phenotyping in genomic prediction

Phenotyping is the current bottleneck in plant breeding, especially because next-generation sequencing has decreased genotyping cost more than 100.000 fold in the last 20 years. Therefore, the cost of phenotyping needs to be optimized within a breeding program. When designing the implementation of genomic selection scheme into the breeding cycle, breeders need to select the optimal method for (1) selecting training populations that maximize genomic prediction accuracy and (2) to reduce the cost of phenotyping while improving precision. In this article, we compared methods for selecting training populations under two scenarios: Firstly, when the objective is to select a training population set (TRS) to predict the remaining individuals from the same population (Untargeted), and secondly, when a test set (TS) is first defined and genotyped, and then the TRS is optimized specifically around the TS (Targeted). Our results show that optimization methods that include information from the test set (targeted) showed the highest accuracies, indicating that apriori information from the TS improves genomic predictions. In addition, predictive ability enhanced especially when population size was small which is a target to decrease phenotypic cost within breeding programs.

Genomic Prediction for 25 Agronomic and Quality Traits in Alfalfa (Medicago sativa)

Agronomic and quality traits in alfalfa are very important to forage industry. Genomic prediction (GP) based on genotyping-by-sequencing (GBS) data could shorten the breeding cycles and accelerate the genetic gains of these complex traits, if they display moderate to high prediction accuracies. The aim of this study was to investigate the predictive potentials of these traits in alfalfa. A total of 322 genotypes from 75 alfalfa accessions were used for GP of the agronomic and quality traits, which were related to yield and nutrition value, respectively, using BayesA, BayesB, and BayesCπ methods. Ten-fold cross validation was used to evaluate the accuracy of GP represented by the correlation between genomic estimated breeding value (GEBV) and estimated breeding value (EBV). The accuracies ranged from 0.0021 to 0.6485 for different traits. For each trait, three GP methods displayed similar prediction accuracies. Among 15 quality traits, mineral element Ca had a moderate and the highest prediction accuracy (0.34). NDF digestibility after 48 h (NDFD 48 h) and 30 h (NDFD 30 h) and mineral element Mg had prediction accuracies varying from 0.20 to 0.25. Other traits, for example, fat and crude protein, showed low prediction accuracies (0.05 to 0.19). Among 10 agronomic traits, however, some displayed relatively high prediction accuracies. Plant height (PH) in fall (FH) had the highest prediction accuracy (0.65), followed by flowering date (FD) and plant regrowth (PR) with accuracies at 0.52 and 0.51, respectively. Leaf to stem ratio (LS), plant branch (PB), and biomass yield (BY) reached to moderate prediction accuracies ranging from 0.25 to 0.32. Our results revealed that a few agronomic traits, such as FH, FD, and PR, had relatively high prediction accuracies, therefore it is feasible to apply genomic selection (GS) for these traits in alfalfa breeding programs. Because of the limitations of population size and density of SNP markers, several traits displayed low accuracies which could be improved by a bigger reference population, higher density of SNP markers, and more powerful statistic tools.

Genomic Selection in Preliminary Yield Trials in a Winter Wheat Breeding Program

Genomic prediction (GP) is now routinely performed in crop plants to predict unobserved phenotypes. The use of predicted phenotypes to make selections is an active area of research. Here, we evaluate GP for predicting grain yield and compare genomic and phenotypic selection by tracking lines advanced. We examined four independent nurseries of F3:6 and F3:7 lines trialed at 6 to 10 locations each year. Yield was analyzed using mixed models that accounted for experimental design and spatial variations. Genotype-by-sequencing provided nearly 27,000 high-quality SNPs. Average genomic predictive ability, estimated for each year by randomly masking lines as missing in steps of 10% from 10 to 90%, and using the remaining lines from the same year as well as lines from other years in a training set, ranged from 0.23 to 0.55. The predictive ability estimated for a new year using the other years ranged from 0.17 to 0.28. Further, we tracked lines advanced based on phenotype from each of the four F3:6 nurseries. Lines with both above average genomic estimated breeding value (GEBV) and phenotypic value (BLUP) were retained for more years compared to lines with either above average GEBV or BLUP alone. The number of lines selected for advancement was substantially greater when predictions were made with 50% of the lines from the testing year added to the training set. Hence, evaluation of only 50% of the lines yearly seems possible. This study provides insights to assess and integrate genomic selection in breeding programs of autogamous crops.

Accounting for Genotype-by-Environment Interactions and Residual Genetic Variation in Genomic Selection for Water-Soluble Carbohydrate Concentration in Wheat

Abiotic stress tolerance traits are often complex and recalcitrant targets for conventional breeding improvement in many crop species. This study evaluated the potential of genomic selection to predict water-soluble carbohydrate concentration (WSCC), an important drought tolerance trait, in wheat under field conditions. A panel of 358 varieties and breeding lines constrained for maturity was evaluated under rainfed and irrigated treatments across two locations and two years. Whole-genome marker profiles and factor analytic mixed models were used to generate genomic estimated breeding values (GEBVs) for specific environments and environment groups. Additive genetic variance was smaller than residual genetic variance for WSCC, such that genotypic values were dominated by residual genetic effects rather than additive breeding values. As a result, GEBVs were not accurate predictors of genotypic values of the extant lines, but GEBVs should be reliable selection criteria to choose parents for intermating to produce new populations. The accuracy of GEBVs for untested lines was sufficient to increase predicted genetic gain from genomic selection per unit time compared to phenotypic selection if the breeding cycle is reduced by half by the use of GEBVs in off-season generations. Further, genomic prediction accuracy depended on having phenotypic data from environments with strong correlations with target production environments to build prediction models. By combining high-density marker genotypes, stress-managed field evaluations, and mixed models that model simultaneously covariances among genotypes and covariances of complex trait performance between pairs of environments, we were able to train models with good accuracy to facilitate genetic gain from genomic selection.

Translating High-Throughput Phenotyping into Genetic Gain

Inability to efficiently implement high-throughput field phenotyping is increasingly perceived as a key component that limits genetic gain in breeding programs. Field phenotyping must be integrated into a wider context than just choosing the correct selection traits, deployment tools, evaluation platforms, or basic data-management methods. Phenotyping means more than conducting such activities in a resource-efficient manner; it also requires appropriate trial management and spatial variability handling, definition of key constraining conditions prevalent in the target population of environments, and the development of more comprehensive data management, including crop modeling. This review will provide a wide perspective on how field phenotyping is best implemented. It will also outline how to bridge the gap between breeders and 'phenotypers' in an effective manner.

Increased genomic prediction accuracy in wheat breeding using a large Australian panel

KEY MESSAGE

Genomic prediction accuracy within a large panel was found to be substantially higher than that previously observed in smaller populations, and also higher than QTL-based prediction. In recent years, genomic selection for wheat breeding has been widely studied, but this has typically been restricted to population sizes under 1000 individuals. To assess its efficacy in germplasm representative of commercial breeding programmes, we used a panel of 10,375 Australian wheat breeding lines to investigate the accuracy of genomic prediction for grain yield, physical grain quality and other physiological traits. To achieve this, the complete panel was phenotyped in a dedicated field trial and genotyped using a custom AxiomTM Affymetrix SNP array. A high-quality consensus map was also constructed, allowing the linkage disequilibrium present in the germplasm to be investigated. Using the complete SNP array, genomic prediction accuracies were found to be substantially higher than those previously observed in smaller populations and also more accurate compared to prediction approaches using a finite number of selected quantitative trait loci. Multi-trait genetic correlations were also assessed at an additive and residual genetic level, identifying a negative genetic correlation between grain yield and protein as well as a positive genetic correlation between grain size and test weight.

Accelerating wheat breeding for end-use quality with multi-trait genomic predictions incorporating near infrared and nuclear magnetic resonance-derived phenotypes

Using NIR and NMR predictions of quality traits overcomes a major barrier for the application of genomic selection to accelerate improvement in grain end-use quality traits of wheat. Grain end-use quality traits are among the most important in wheat breeding. These traits are difficult to breed for, as their assays require flour quantities only obtainable late in the breeding cycle, and are expensive. These traits are therefore an ideal target for genomic selection. However, large reference populations are required for accurate genomic predictions, which are challenging to assemble for these traits for the same reasons they are challenging to breed for. Here, we use predictions of end-use quality derived from near infrared (NIR) or nuclear magnetic resonance (NMR), that require very small amounts of flour, as well as end-use quality measured by industry standard assay in a subset of accessions, in a multi-trait approach for genomic prediction. The NIR and NMR predictions were derived for 19 end-use quality traits in 398 accessions, and were then assayed in 2420 diverse wheat accessions. The accessions were grown out in multiple locations and multiple years, and were genotyped for 51208 SNP. Incorporating NIR and NMR phenotypes in the multi-trait approach increased the accuracy of genomic prediction for most quality traits. The accuracy ranged from 0 to 0.47 before the addition of the NIR/NMR data, while after these data were added, it ranged from 0 to 0.69. Genomic predictions were reasonably robust across locations and years for most traits. Using NIR and NMR predictions of quality traits overcomes a major barrier for the application of genomic selection for grain end-use quality traits in wheat breeding.

Modelling spatial trends in sorghum breeding field trials using a two-dimensional P-spline mixed model

A flexible and user-friendly spatial method called SpATS performed comparably to more elaborate and trial-specific spatial models in a series of sorghum breeding trials. Adjustment for spatial trends in plant breeding field trials is essential for efficient evaluation and selection of genotypes. Current mixed model methods of spatial analysis are based on a multi-step modelling process where global and local trends are fitted after trying several candidate spatial models. This paper reports the application of a novel spatial method that accounts for all types of continuous field variation in a single modelling step by fitting a smooth surface. The method uses two-dimensional P-splines with anisotropic smoothing formulated in the mixed model framework, referred to as SpATS model. We applied this methodology to a series of large and partially replicated sorghum breeding trials. The new model was assessed in comparison with the more elaborate standard spatial models that use autoregressive correlation of residuals. The improvements in precision and the predictions of genotypic values produced by the SpATS model were equivalent to those obtained using the best fitting standard spatial models for each trial. One advantage of the approach with SpATS is that all patterns of spatial trend and genetic effects were modelled simultaneously by fitting a single model. Furthermore, we used a flexible model to adequately adjust for field trends. This strategy reduces potential parameter identification problems and simplifies the model selection process. Therefore, the new method should be considered as an efficient and easy-to-use alternative for routine analyses of plant breeding trials.

Genetic Architecture of Phenomic-Enabled Canopy Coverage in Glycine max

Digital imagery can help to quantify seasonal changes in desirable crop phenotypes that can be treated as quantitative traits. Because limitations in precise and functional phenotyping restrain genetic improvement in the postgenomic era, imagery-based phenomics could become the next breakthrough to accelerate genetic gains in field crops. Whereas many phenomic studies focus on exploratory analysis of spectral data without obvious interpretative value, we used field images to directly measure soybean canopy development from phenological stage V2 to R5. Over 3 years, we collected imagery using ground and aerial platforms of a large and diverse nested association panel comprising 5555 lines. Genome-wide association analysis of canopy coverage across sampling dates detected a large quantitative trait locus (QTL) on soybean (Glycine max, L. Merr.) chromosome 19. This QTL provided an increase in yield of 47.3 kg ha-1 Variance component analysis indicated that a parameter, described as average canopy coverage, is a highly heritable trait (h2 = 0.77) with a promising genetic correlation with grain yield (0.87), enabling indirect selection of yield via canopy development parameters. Our findings indicate that fast canopy coverage is an early season trait that is inexpensive to measure and has great potential for application in breeding programs focused on yield improvement. We recommend using the average canopy coverage in multiple trait schemes, especially for the early stages of the breeding pipeline (including progeny rows and preliminary yield trials), in which the large number of field plots makes collection of grain yield data challenging.

Genome-wide association mapping in winter barley for grain yield and culm cell wall polymer content using the high-throughput CoMPP technique

A collection of 112 winter barley varieties (Hordeum vulgare L.) was grown in the field for two years (2008/09 and 2009/10) in northern Italy and grain and straw yields recorded. In the first year of the trial, a severe attack of barley yellow mosaic virus (BaYMV) strongly influenced final performances with an average reduction of ~ 50% for grain and straw harvested in comparison to the second year. The genetic determination (GD) for grain yield was 0.49 and 0.70, for the two years respectively, and for straw yield GD was low in 2009 (0.09) and higher in 2010 (0.29). Cell wall polymers in culms were quantified by means of the monoclonal antibodies LM6, LM11, JIM13 and BS-400-3 and the carbohydrate-binding module CBM3a using the high-throughput CoMPP technique. Of these, LM6, which detects arabinan components, showed a relatively high GD in both years and a significantly negative correlation with grain yield (GYLD). Overall, heritability (H2) was calculated for GYLD, LM6 and JIM and resulted to be 0.42, 0.32 and 0.20, respectively. A total of 4,976 SNPs from the 9K iSelect array were used in the study for the analysis of population structure, linkage disequilibrium (LD) and genome-wide association study (GWAS). Marker-trait associations (MTA) were analyzed for grain yield and cell wall determination by LM6 and JIM13 as these were the traits showing significant correlations between the years. A single QTL for GYLD containing three MTAs was found on chromosome 3H located close to the Hv-eIF4E gene, which is known to regulate resistance to BaYMV. Subsequently the QTL was shown to be tightly linked to rym4, a locus for resistance to the virus. GWAs on arabinans quantified by LM6 resulted in the identification of major QTLs closely located on 3H and hypotheses regarding putative candidate genes were formulated through the study of gene expression levels based on bioinformatics tools.

Accuracy of Genomic Prediction in a Commercial Perennial Ryegrass Breeding Program

The implementation of genomic selection (GS) in plant breeding, so far, has been mainly evaluated in crops farmed as homogeneous varieties, and the results have been generally positive. Fewer results are available for species, such as forage grasses, that are grown as heterogenous families (developed from multiparent crosses) in which the control of the genetic variation is far more complex. Here we test the potential for implementing GS in the breeding of perennial ryegrass ( L.) using empirical data from a commercial forage breeding program. Biparental F and multiparental synthetic (SYN) families of diploid perennial ryegrass were genotyped using genotyping-by-sequencing, and phenotypes for five different traits were analyzed. Genotypes were expressed as family allele frequencies, and phenotypes were recorded as family means. Different models for genomic prediction were compared by using practically relevant cross-validation strategies. All traits showed a highly significant level of genetic variance, which could be traced using the genotyping assay. While there was significant genotype × environment (G × E) interaction for some traits, accuracies were high among F families and between biparental F and multiparental SYN families. We have demonstrated that the implementation of GS in grass breeding is now possible and presents an opportunity to make significant gains for various traits.

Ascertainment bias from imputation methods evaluation in wheat

BACKGROUND

Whole-genome genotyping techniques like Genotyping-by-sequencing (GBS) are being used for genetic studies such as Genome-Wide Association (GWAS) and Genomewide Selection (GS), where different strategies for imputation have been developed. Nevertheless, imputation error may lead to poor performance (i.e. smaller power or higher false positive rate) when complete data is not required as it is for GWAS, and each marker is taken at a time. The aim of this study was to compare the performance of GWAS analysis for Quantitative Trait Loci (QTL) of major and minor effect using different imputation methods when no reference panel is available in a wheat GBS panel.

RESULTS

In this study, we compared the power and false positive rate of dissecting quantitative traits for imputed and not-imputed marker score matrices in: (1) a complete molecular marker barley panel array, and (2) a GBS wheat panel with missing data. We found that there is an ascertainment bias in imputation method comparisons. Simulating over a complete matrix and creating missing data at random proved that imputation methods have a poorer performance. Furthermore, we found that when QTL were simulated with imputed data, the imputation methods performed better than the not-imputed ones. On the other hand, when QTL were simulated with not-imputed data, the not-imputed method and one of the imputation methods performed better for dissecting quantitative traits. Moreover, larger differences between imputation methods were detected for QTL of major effect than QTL of minor effect. We also compared the different marker score matrices for GWAS analysis in a real wheat phenotype dataset, and we found minimal differences indicating that imputation did not improve the GWAS performance when a reference panel was not available.

CONCLUSIONS

Poorer performance was found in GWAS analysis when an imputed marker score matrix was used, no reference panel is available, in a wheat GBS panel.

Walking through the statistical black boxes of plant breeding

The main statistical procedures in plant breeding are based on Gaussian process and can be computed through mixed linear models. Intelligent decision making relies on our ability to extract useful information from data to help us achieve our goals more efficiently. Many plant breeders and geneticists perform statistical analyses without understanding the underlying assumptions of the methods or their strengths and pitfalls. In other words, they treat these statistical methods (software and programs) like black boxes. Black boxes represent complex pieces of machinery with contents that are not fully understood by the user. The user sees the inputs and outputs without knowing how the outputs are generated. By providing a general background on statistical methodologies, this review aims (1) to introduce basic concepts of machine learning and its applications to plant breeding; (2) to link classical selection theory to current statistical approaches; (3) to show how to solve mixed models and extend their application to pedigree-based and genomic-based prediction; and (4) to clarify how the algorithms of genome-wide association studies work, including their assumptions and limitations.

Optimizing Training Population Data and Validation of Genomic Selection for Economic Traits in Soft Winter Wheat

Genomic selection (GS) is a breeding tool that estimates breeding values (GEBVs) of individuals based solely on marker data by using a model built using phenotypic and marker data from a training population (TP). The effectiveness of GS increases as the correlation of GEBVs and phenotypes (accuracy) increases. Using phenotypic and genotypic data from a TP of 470 soft winter wheat lines, we assessed the accuracy of GS for grain yield, Fusarium Head Blight (FHB) resistance, softness equivalence (SE), and flour yield (FY). Four TP data sampling schemes were tested: (1) use all TP data, (2) use subsets of TP lines with low genotype-by-environment interaction, (3) use subsets of markers significantly associated with quantitative trait loci (QTL), and (4) a combination of 2 and 3. We also correlated the phenotypes of relatives of the TP to their GEBVs calculated from TP data. The GS accuracy within the TP using all TP data ranged from 0.35 (FHB) to 0.62 (FY). On average, the accuracy of GS from using subsets of data increased by 54% relative to using all TP data. Using subsets of markers selected for significant association with the target trait had the greatest impact on GS accuracy. Between-environment prediction accuracy was also increased by using data subsets. The accuracy of GS when predicting the phenotypes of TP relatives ranged from 0.00 to 0.85. These results suggest that GS could be useful for these traits and GS accuracy can be greatly improved by using subsets of TP data.

Genomic prediction for grain zinc and iron concentrations in spring wheat

Predictability estimated through cross-validation approach showed moderate to high level; hence, genomic selection approach holds great potential for biofortification breeding to enhance grain zinc and iron concentrations in wheat. Wheat (Triticum aestivum L.) is a major staple crop, providing 20 % of dietary energy and protein consumption worldwide. It is an important source of mineral micronutrients such as zinc (Zn) and iron (Fe) for resource poor consumers. Genomic selection (GS) approaches have great potential to accelerate development of Fe- and Zn-enriched wheat. Here, we present the results of large-scale genomic and phenotypic data from the HarvestPlus Association Mapping (HPAM) panel consisting of 330 diverse wheat lines to perform genomic predictions for grain Zn (GZnC) and Fe (GFeC) concentrations, thousand-kernel weight (TKW) and days to maturity (DTM) in wheat. The HPAM lines were phenotyped in three different locations in India and Mexico in two successive crop seasons (2011-12 and 2012-13) for GZnC, GFeC, TKW and DTM. The genomic prediction models revealed that the estimated prediction abilities ranged from 0.331 to 0.694 for Zn and from 0.324 to 0.734 for Fe according to different environments, whereas prediction abilities for TKW and DTM were as high as 0.76 and 0.64, respectively, suggesting that GS holds great potential in biofortification breeding to enhance grain Zn and Fe concentrations in bread wheat germplasm.

Genomic Prediction of Manganese Efficiency in Winter Barley

Manganese efficiency is a quantitative abiotic stress trait controlled by several genes each with a small effect. Manganese deficiency leads to yield reduction in winter barley ( L.). Breeding new cultivars for this trait remains difficult because of the lack of visual symptoms and the polygenic features of the trait. Hence, Mn efficiency is a potential suitable trait for a genomic selection (GS) approach. A collection of 248 winter barley varieties was screened for Mn efficiency using Chlorophyll (Chl ) fluorescence in six environments prone to induce Mn deficiency. Two models for genomic prediction were implemented to predict future performance and breeding value of untested varieties. Predictions were obtained using multivariate mixed models: best linear unbiased predictor (BLUP) and genomic best linear unbiased predictor (G-BLUP). In the first model, predictions were based on the phenotypic evaluation, whereas both phenotypic and genomic marker data were included in the second model. Accuracy of predicting future phenotype, , and accuracy of predicting true breeding values, , were calculated and compared for both models using six cross-validation (CV) schemes; these were designed to mimic plant breeding programs. Overall, the CVs showed that prediction accuracies increased when using the G-BLUP model compared with the prediction accuracies using the BLUP model. Furthermore, the accuracies [] of predicting breeding values were more accurate than accuracy of predicting future phenotypes []. The study confirms that genomic data may enhance the prediction accuracy. Moreover it indicates that GS is a suitable breeding approach for quantitative abiotic stress traits.

Genomic Selection in Multi-environment Crop Trials

Genomic selection in crop breeding introduces modeling challenges not found in animal studies. These include the need to accommodate replicate plants for each line, consider spatial variation in field trials, address line by environment interactions, and capture nonadditive effects. Here, we propose a flexible single-stage genomic selection approach that resolves these issues. Our linear mixed model incorporates spatial variation through environment-specific terms, and also randomization-based design terms. It considers marker, and marker by environment interactions using ridge regression best linear unbiased prediction to extend genomic selection to multiple environments. Since the approach uses the raw data from line replicates, the line genetic variation is partitioned into marker and nonmarker residual genetic variation (i.e., additive and nonadditive effects). This results in a more precise estimate of marker genetic effects. Using barley height data from trials, in 2 different years, of up to 477 cultivars, we demonstrate that our new genomic selection model improves predictions compared to current models. Analyzing single trials revealed improvements in predictive ability of up to 5.7%. For the multiple environment trial (MET) model, combining both year trials improved predictive ability up to 11.4% compared to a single environment analysis. Benefits were significant even when fewer markers were used. Compared to a single-year standard model run with 3490 markers, our partitioned MET model achieved the same predictive ability using between 500 and 1000 markers depending on the trial. Our approach can be used to increase accuracy and confidence in the selection of the best lines for breeding and/or, to reduce costs by using fewer markers.

Genomic Selection in Multi-environment Crop Trials

Genomic selection in crop breeding introduces modeling challenges not found in animal studies. These include the need to accommodate replicate plants for each line, consider spatial variation in field trials, address line by environment interactions, and capture nonadditive effects. Here, we propose a flexible single-stage genomic selection approach that resolves these issues. Our linear mixed model incorporates spatial variation through environment-specific terms, and also randomization-based design terms. It considers marker, and marker by environment interactions using ridge regression best linear unbiased prediction to extend genomic selection to multiple environments. Since the approach uses the raw data from line replicates, the line genetic variation is partitioned into marker and nonmarker residual genetic variation (i.e., additive and nonadditive effects). This results in a more precise estimate of marker genetic effects. Using barley height data from trials, in 2 different years, of up to 477 cultivars, we demonstrate that our new genomic selection model improves predictions compared to current models. Analyzing single trials revealed improvements in predictive ability of up to 5.7%. For the multiple environment trial (MET) model, combining both year trials improved predictive ability up to 11.4% compared to a single environment analysis. Benefits were significant even when fewer markers were used. Compared to a single-year standard model run with 3490 markers, our partitioned MET model achieved the same predictive ability using between 500 and 1000 markers depending on the trial. Our approach can be used to increase accuracy and confidence in the selection of the best lines for breeding and/or, to reduce costs by using fewer markers.

Comparing the Predictive Abilities of Phenotypic and Marker-Assisted Selection Methods in a Biparental Lettuce Population

Breeding for traits with polygenic inheritance is a challenging task that can be done by phenotypic selection, marker-assisted selection (MAS) or genome-wide selection. We comparatively evaluated the predictive abilities of four selection models on a biparental lettuce ( L.) population genotyped with 95 single nucleotide polymorphisms and 205 amplified fragment length polymorphism markers. These models were based on (i) phenotypic selection, (ii) MAS (with quantitative trait locus (QTL)-linked markers), (iii) genomic prediction using all the available molecular markers, and (iv) genomic prediction using molecular markers plus QTL-linked markers as fixed covariates. Each model's performance was assessed using data on the field resistance to downy mildew (DMR, mean heritability ∼0.71) and the quality of shelf life (SL, mean heritability ∼0.91) of lettuce in multiple environments. The predictive ability of each selection model was computed under three cross-validation (CV) schemes based on sampling genotypes, environments, or both. For the DMR dataset, the predictive ability of the MAS model was significantly lower than that of the genomic prediction model. For the SL dataset, the predictive ability of the genomic prediction model was significantly lower than that for the model using QTL-linked markers under two of the three CV schemes. Our results show that the predictive ability of the selection models depends strongly on the CV scheme used for prediction and the heritability of the target trait. Our study also shows that molecular markers can be used to predict DMR and SL for individuals from this cross that were genotyped but not phenotyped.

Latin America: A Development Pole for Phenomics

Latin America and the Caribbean (LAC) has long been associated with the production and export of a diverse range of agricultural commodities. Due to its strategic geographic location, which encompasses a wide range of climates, it is possible to produce almost any crop. The climate diversity in LAC is a major factor in its agricultural potential but this also means climate change represents a real threat to the region. Therefore, LAC farming must prepare and quickly adapt to an environment that is likely to feature long periods of drought, excessive rainfall and extreme temperatures. With the aim of moving toward a more resilient agriculture, LAC scientists have created the Latin American Plant Phenomics Network (LatPPN) which focuses on LAC's economically important crops. LatPPN's key strategies to achieve its main goal are: (1) training of LAC members on plant phenomics and phenotyping, (2) establish international and multidisciplinary collaborations, (3) develop standards for data exchange and research protocols, (4) share equipment and infrastructure, (5) disseminate data and research results, (6) identify funding opportunities and (7) develop strategies to guarantee LatPPN's relevance and sustainability across time. Despite the challenges ahead, LatPPN represents a big step forward toward the consolidation of a common mind-set in the field of plant phenotyping and phenomics in LAC.

Next generation breeding

The genomic revolution of the past decade has greatly improved our understanding of the genetic make-up of living organisms. The sequencing of crop genomes has completely changed our vision and interpretation of genome organization and evolution. Re-sequencing allows the identification of an unlimited number of markers as well as the analysis of germplasm allelic diversity based on allele mining approaches. High throughput marker technologies coupled with advanced phenotyping platforms provide new opportunities for discovering marker-trait associations which can sustain genomic-assisted breeding. The availability of genome sequencing information is enabling genome editing (site-specific mutagenesis), to obtain gene sequences desired by breeders. This review illustrates how next generation sequencing-derived information can be used to tailor genomic tools for different breeders' needs to revolutionize crop improvement.

Association Mapping in Scandinavian Winter Wheat for Yield, Plant Height, and Traits Important for Second-Generation Bioethanol Production

A collection of 100 wheat varieties representing more than 100 years of wheat-breeding history in Scandinavia was established in order to identify marker-trait associations for plant height (PH), grain yield (GY), and biomass potential for bioethanol production. The field-grown material showed variations in PH from 54 to 122 cm and in GY from 2 to 6.61 t ha(-1). The release of monomeric sugars was determined by high-throughput enzymatic treatment of ligno-cellulosic material and varied between 0.169 and 0.312 g/g dm for glucose (GLU) and 0.146 and 0.283 g/g dm for xylose (XYL). As expected, PH and GY showed to be highly influenced by genetic factors with repeatability (R) equal to 0.75 and 0.53, respectively, while this was reduced for GLU and XYL (R = 0.09 for both). The study of trait correlations showed how old, low-yielding, tall varieties released higher amounts of monomeric sugars after straw enzymatic hydrolysis, showing reduced recalcitrance to bioconversion compared to modern varieties. Ninety-three lines from the collection were genotyped with the DArTseq(®) genotypic platform and 5525 markers were used for genome-wide association mapping. Six quantitative trait loci (QTLs) for GY, PH, and GLU released from straw were mapped. One QTL for PH was previously reported, while the remaining QTLs constituted new genomic regions linked to trait variation. This paper is one of the first studies in wheat to identify QTLs that are important for bioethanol production based on a genome-wide association approach.

Genomic Selection for Predicting Fusarium Head Blight Resistance in a Wheat Breeding Program

Genomic selection (GS) is a breeding method that uses marker-trait models to predict unobserved phenotypes. This study developed GS models for predicting traits associated with resistance to Fusarium head blight (FHB) in wheat (Triticum aestivum L.). We used genotyping-by-sequencing (GBS) to identify 5054 single-nucleotide polymorphisms (SNPs), which were then treated as predictor variables in GS analysis. We compared how the prediction accuracy of the genomic-estimated breeding values (GEBVs) was affected by (i) five genotypic imputation methods (random forest imputation [RFI], expectation maximization imputation [EMI], k-nearest neighbor imputation [kNNI], singular value decomposition imputation [SVDI], and the mean imputation [MNI]); (ii) three statistical models (ridge-regression best linear unbiased predictor [RR-BLUP], least absolute shrinkage and operator selector [LASSO], and elastic net); (iii) marker density (p = 500, 1500, 3000, and 4500 SNPs); (iv) training population (TP) size (n_TP = 96, 144, 192, and 218); (v) marker-based and pedigree-based relationship matrices; and (vi) control for relatedness in TPs and validation populations (VPs). No discernable differences in prediction accuracy were observed among imputation methods. The RR-BLUP outperformed other models in nearly all scenarios. Accuracies decreased substantially when marker number decreased to 3000 or 1500 SNPs, depending on the trait; when sample size of the training set was less than 192; when using pedigree-based instead of marker-based matrix; or when no control for relatedness was implemented. Overall, moderate to high prediction accuracies were observed in this study, suggesting that GS is a very promising breeding strategy for FHB resistance in wheat.

Breaking through the feed barrier: options for improving forage genetics

Pasture based on perennial ryegrass (Lolium perenne L.) and white clover (Trifolium repens L.) is the foundation for production and profit in the Australasian pastoral sectors. The improvement of these species offers direct opportunities to enhance sector performance, provided there is good alignment with industry priorities as quantified by means such as the forage value index. However, the rate of forage genetic improvement must increase to sustain industry competitiveness. New forage technologies and breeding strategies that can complement and enhance traditional approaches are required to achieve this. We highlight current and future research in plant breeding, including genomic and gene technology approaches to improve rate of genetic gain. Genomic diversity is the basis of breeding and improvement. Recent advances in the range and focus of introgression from wild Trifolium species have created additional specific options to improve production and resource-use-efficiency traits. Symbiont genetic resources, especially advances in grass fungal endophytes, make a critical contribution to forage, supporting pastoral productivity, with benefits to both pastures and animals in some dairy regions. Genomic selection, now widely used in animal breeding, offers an opportunity to lift the rate of genetic gain in forages as well. Accuracy and relevance of trait data are paramount, it is essential that genomic breeding approaches be linked with robust field evaluation strategies including advanced phenotyping technologies. This requires excellent data management and integration with decision-support systems to deliver improved effectiveness from forage breeding. Novel traits being developed through genetic modification include increased energy content and potential increased biomass in ryegrass, and expression of condensed tannins in forage legumes. These examples from the wider set of research emphasise forage adaptation, yield and energy content, while covering the spectrum from exotic germplasm and symbionts through to advanced breeding strategies and gene technologies. To ensure that these opportunities are realised on farm, continuity of industry-relevant delivery of forage-improvement research is essential, as is sustained research input from the supporting pasture and plant sciences.