Unlocking big data doubled the accuracy in predicting the grain yield in hybrid wheat

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.

Deciphering the evolution and complexity of wheat germplasm from a genomic perspective

Bread wheat provides an essential fraction of the daily calorific intake for humanity. Due to its huge and complex genome, progress in studying on the wheat genome is substantially trailed behind those of the other two major crops, rice and maize, for at least a decade. With rapid advances in genome assembling and reduced cost of high-throughput sequencing, emerging de novo genome assemblies of wheat and whole-genome sequencing data are leading to a paradigm shift in wheat research. Here, we review recent progress in dissecting the complex genome and germplasm evolution of wheat since the release of the first high-quality wheat genome. New insights have been gained in the evolution of wheat germplasm during domestication and modern breeding progress, genomic variations at multiple scales contributing to the diversity of wheat germplasm, and complex transcriptional and epigenetic regulations of functional genes in polyploid wheat. Genomics databases and bioinformatics tools meeting the urgent needs of wheat genomics research are also summarized. The ever-increasing omics data, along with advanced tools and well-structured databases, are expected to accelerate deciphering the germplasm and gene resources in wheat for future breeding advances.

Stacked ensembles on basis of parentage information can predict hybrid performance with an accuracy comparable to marker-based GBLUP

Testcross factorials in newly established hybrid breeding programs are often highly unbalanced, incomplete, and characterized by predominance of special combining ability (SCA) over general combining ability (GCA). This results in a low efficiency of GCA-based selection. Machine learning algorithms might improve prediction of hybrid performance in such testcross factorials, as they have been successfully applied to find complex underlying patterns in sparse data. Our objective was to compare the prediction accuracy of machine learning algorithms to that of GCA-based prediction and genomic best linear unbiased prediction (GBLUP) in six unbalanced incomplete factorials from hybrid breeding programs of rapeseed, wheat, and corn. We investigated a range of machine learning algorithms with three different types of predictor variables: (a) information on parentage of hybrids, (b) in addition hybrid performance of crosses of the parental lines with other crossing partners, and (c) genotypic marker data. In two highly incomplete and unbalanced factorials from rapeseed, in which the SCA variance contributed considerably to the genetic variance, stacked ensembles of gradient boosting machines based on parentage information outperformed GCA prediction. The stacked ensembles increased prediction accuracy from 0.39 to 0.45, and from 0.48 to 0.54 compared to GCA prediction. The prediction accuracy reached by stacked ensembles without marker data reached values comparable to those of GBLUP that requires marker data. We conclude that hybrid prediction with stacked ensembles of gradient boosting machines based on parentage information is a promising approach that is worth further investigations with other data sets in which SCA variance is high.

Biofortification to avoid malnutrition in humans in a changing climate: Enhancing micronutrient bioavailability in seed, tuber, and storage roots

Malnutrition results in enormous socio-economic costs to the individual, their community, and the nation's economy. The evidence suggests an overall negative impact of climate change on the agricultural productivity and nutritional quality of food crops. Producing more food with better nutritional quality, which is feasible, should be prioritized in crop improvement programs. Biofortification refers to developing micronutrient -dense cultivars through crossbreeding or genetic engineering. This review provides updates on nutrient acquisition, transport, and storage in plant organs; the cross-talk between macro- and micronutrients transport and signaling; nutrient profiling and spatial and temporal distribution; the putative and functionally characterized genes/single-nucleotide polymorphisms associated with Fe, Zn, and β-carotene; and global efforts to breed nutrient-dense crops and map adoption of such crops globally. This article also includes an overview on the bioavailability, bioaccessibility, and bioactivity of nutrients as well as the molecular basis of nutrient transport and absorption in human. Over 400 minerals (Fe, Zn) and provitamin A-rich cultivars have been released in the Global South. Approximately 4.6 million households currently cultivate Zn-rich rice and wheat, while ~3 million households in sub-Saharan Africa and Latin America benefit from Fe-rich beans, and 2.6 million people in sub-Saharan Africa and Brazil eat provitamin A-rich cassava. Furthermore, nutrient profiles can be improved through genetic engineering in an agronomically acceptable genetic background. The development of "Golden Rice" and provitamin A-rich dessert bananas and subsequent transfer of this trait into locally adapted cultivars are evident, with no significant change in nutritional profile, except for the trait incorporated. A greater understanding of nutrient transport and absorption may lead to the development of diet therapy for the betterment of human health.

Population structure and genetic diversity of a germplasm for hybrid breeding in rye (Secale cereale L.) using high-density DArTseq-based silicoDArT and SNP markers

Investigating genetic structure and diversity is crucial for the rye hybrid breeding strategy, leading to improved plant productivity and adaptation. The present study elucidated the population structure and genetic diversity of 188 rye accessions, comprising 94 pollen fertility restoration lines (RF) and 94 cytoplasmic male-sterile (CMS) lines with Pampa sterilizing cytoplasm using SNP and silicoDArT markers from the diversity array technology (DArT)-based sequencing platform (DArTseq). Expected heterozygosity (H_e) and Shanon's diversity (I) indexes varied slightly between marker systems and groups of germplasms (H_e = 0.34, I = 0.51 for RF and CMS lines genotyped using SNPs; H_e = 0.31, I = 0.48, and H_e = 0.35, I = 0.53 for RF and CMS using silicoDArTs, respectively). ANOVA indicated moderate variation (7%) between RF and CMS breeding materials. The same parameter varied when chromosome-assigned markers were used and ranged from 5.8% for 5R to 7.4% for 4R. However, when silicoDArT markers were applied, the respective values varied from 6.4% (1R) to 8.2% (3R and 4R). The model-based (Bayesian) population structure analysis based on the total marker pool identified two major subpopulations for the studied rye germplasm. The first one (P1) encompasses 93 RF accessions, and the second one (P2) encompasses 94 CMS and one RF accession. However, a similar analysis related to markers assigned to selected chromosomes failed to put plant materials into any of the populations in the same way as the total marker pool. Furthermore, the differences in grouping depended on marker types used for analysis.

Imputation of low-density marker chip data in plant breeding: Evaluation of methods based on sugar beet

Low-density genotyping followed by imputation reduces genotyping costs while still providing high-density marker information. An increased marker density has the potential to improve the outcome of all applications that are based on genomic data. This study investigates techniques for 1k to 20k genomic marker imputation for plant breeding programs with sugar beet (Beta vulgaris L. ssp. vulgaris) as an example crop, where these are realistic marker numbers for modern breeding applications. The generally accepted 'gold standard' for imputation, Beagle 5.1, was compared with the recently developed software AlphaPlantImpute2 which is designed specifically for plant breeding. For Beagle 5.1 and AlphaPlantImpute2, the imputation strategy as well as the imputation parameters were optimized in this study. We found that the imputation accuracy of Beagle could be tremendously improved (0.22 to 0.67) by tuning parameters, mainly by lowering the values for the parameter for the effective population size and increasing the number of iterations performed. Separating the phasing and imputation steps also improved accuracies when optimized parameters were used (0.67 to 0.82). We also found that the imputation accuracy of Beagle decreased when more low-density lines were included for imputation. AlphaPlantImpute2 produced very high accuracies without optimization (0.89) and was generally less responsive to optimization. Overall, AlphaPlantImpute2 performed relatively better for imputation whereas Beagle was better for phasing. Combining both tools yielded the highest accuracies.

Utilizing genomics and historical data to optimize gene pools for new breeding programs: A case study in winter wheat

With the rapid generation and preservation of both genomic and phenotypic information for many genotypes within crops and across locations, emerging breeding programs have a valuable opportunity to leverage these resources to 1) establish the most appropriate genetic foundation at program inception and 2) implement robust genomic prediction platforms that can effectively select future breeding lines. Integrating genomics-enabled¹ breeding into cultivar development can save costs and allow resources to be reallocated towards advanced (i.e., later) stages of field evaluation, which can facilitate an increased number of testing locations and replicates within locations. In this context, a reestablished winter wheat breeding program was used as a case study to understand best practices to leverage and tailor existing genomic and phenotypic resources to determine optimal genetics for a specific target population of environments. First, historical multi-environment phenotype data, representing 1,285 advanced breeding lines, were compiled from multi-institutional testing as part of the SunGrains cooperative and used to produce GGE biplots and PCA for yield. Locations were clustered based on highly correlated line performance among the target population of environments into 22 subsets. For each of the subsets generated, EMMs and BLUPs were calculated using linear models with the ‘lme4’ R package. Second, for each subset, TPs representative of the new SC breeding lines were determined based on genetic relatedness using the ‘STPGA’ R package. Third, for each TP, phenotypic values and SNP data were incorporated into the ‘rrBLUP’ mixed models for generation of GEBVs of YLD, TW, HD and PH. Using a five-fold cross-validation strategy, an average accuracy of r = 0.42 was obtained for yield between all TPs. The validation performed with 58 SC elite breeding lines resulted in an accuracy of r = 0.62 when the TP included complete historical data. Lastly, QTL-by-environment interaction for 18 major effect genes across three geographic regions was examined. Lines harboring major QTL in the absence of disease could potentially underperform (e.g., Fhb1 R-gene), whereas it is advantageous to express a major QTL under biotic pressure (e.g., stripe rust R-gene). This study highlights the importance of genomics-enabled breeding and multi-institutional partnerships to accelerate cultivar development.

Crucial factors for the feasibility of commercial hybrid breeding in food crops

There is an ongoing societal debate about plant breeding systems and their impact on stakeholders in food systems. Hybrid breeding and hybrid seed have become controversial topics as they are believed to mostly serve high-tech agricultural systems. This article focuses on the perspective of commercial plant breeders when developing new cultivars of food crops. Arguably, hybrid breeding is the most effective breeding system for genetic improvement of crops, enhancing yields, improving product quality and increasing resistance against (a)biotic stresses. Nonetheless, hybrid breeding is not commercially applied in all crops. We analyse how biological and economic factors determine whether a commercial plant breeder opts for the hybrid system or not. We show that the commercial feasibility of hybrid breeding depends on the crop and business case. In conclusion, the commercial application of hybrid breeding in crops seems to be hampered mostly by high costs of seed production. Case studies regarding the hybrid transitions in maize, wheat and potato are included to illustrate these findings.

Hybrid wheat: quantitative genetic parameters and heterosis for quality and rheological traits as well as baking volume

Heterosis effects for dough quality and baking volume were close to zero. However, hybrids have a higher grain yield at a given level of bread making quality compared to their parental lines. Bread wheat cultivars have been selected according to numerous quality traits to fulfill the requirements of the bread making industry. These include beside protein content and quality also rheological traits and baking volume. We evaluated 35 male and 73 female lines and 119 of their single-cross hybrids at three different locations for grain yield, protein content, sedimentation value, extensograph traits and baking volume. No significant differences (p < 0.05) were found in the mean comparisons of males, females and hybrids, except for higher grain yield and lower protein content in the hybrids. Mid-parent and better-parent heterosis values were close to zero and slightly negative, respectively, for baking volume and extensograph traits. However, the majority of heterosis values resulted in the finding that hybrids had higher grain yield than lines for a given level of baking volume, sedimentation value or energy value of extensograph. Due to the high correlation with the mid-parent values (r > 0.70), an initial prediction of hybrid performance based on line per se performance for protein content, sedimentation value, most traits of the extensograph and baking volume is possible. The low variance due to specific combining ability effects for most quality traits points toward an additive gene action requires quality selection within both heterotic groups. Consequently, hybrid wheat can combine high grain yield with high bread making quality. However, the future use of wheat hybrids strongly depends on the establishment of a cost-efficient and reliable seed production system.

Big genomic data analysis leads to more accurate trait prediction in hybrid breeding for yield enhancement in crop plants

The 'big data' in plant breeding refers to the cumulative genotyping and phenotyping information obtained from either a series of experimental sets or generated from a large number of accessions. Recent study supports the employment of big data for enhancing the accuracy of complex trait prediction during hybrid breeding of crop plants.

Reciprocal Recurrent Genomic Selection Is Impacted by Genotype-by-Environment Interactions

Reciprocal recurrent genomic selection is a breeding strategy aimed at improving the hybrid performance of two base populations. It promises to significantly advance hybrid breeding in wheat. Against this backdrop, the main objective of this study was to empirically investigate the potential and limitations of reciprocal recurrent genomic selection. Genome-wide predictive equations were developed using genomic and phenotypic data from a comprehensive population of 1,604 single crosses between 120 female and 15 male wheat lines. Twenty superior female lines were selected for initiation of the reciprocal recurrent genomic selection program. Focusing on the female pool, one cycle was performed with genomic selection steps at the F₂ (60 out of 629 plants) and the F₅ stage (49 out of 382 plants). Selection gain for grain yield was evaluated at six locations. Analyses of the phenotypic data showed pronounced genotype-by-environment interactions with two environments that formed an outgroup compared to the environments used for the genome-wide prediction equations. Removing these two environments for further analysis resulted in a selection gain of 1.0 dt ha^-1 compared to the hybrids of the original 20 parental lines. This underscores the potential of reciprocal recurrent genomic selection to promote hybrid wheat breeding, but also highlights the need to develop robust genome-wide predictive equations.

Breeding custom-designed crops for improved drought adaptation

The current pace of crop improvement is inadequate to feed the burgeoning human population by 2050. Higher, more stable, and sustainable crop production is required against a backdrop of drought stress, which causes significant losses in crop yields. Tailoring crops for drought adaptation may hold the key to address these challenges and provide resilient production systems for future harvests. Understanding the genetic and molecular landscape of the functionality of alleles associated with adaptive traits will make designer crop breeding the prospective approach for crop improvement. Here, we highlight the potential of genomics technologies combined with crop physiology for high-throughput identification of the genetic architecture of key drought-adaptive traits and explore innovative genomic breeding strategies for designing future crops.

Training Set Optimization for Sparse Phenotyping in Genomic Selection: A Conceptual Overview

Genomic selection (GS) is becoming an essential tool in breeding programs due to its role in increasing genetic gain per unit time. The design of the training set (TRS) in GS is one of the key steps in the implementation of GS in plant and animal breeding programs mainly because (i) TRS optimization is critical for the efficiency and effectiveness of GS, (ii) breeders test genotypes in multi-year and multi-location trials to select the best-performing ones. In this framework, TRS optimization can help to decrease the number of genotypes to be tested and, therefore, reduce phenotyping cost and time, and (iii) we can obtain better prediction accuracies from optimally selected TRS than an arbitrary TRS. Here, we concentrate the efforts on reviewing the lessons learned from TRS optimization studies and their impact on crop breeding and discuss important features for the success of TRS optimization under different scenarios. In this article, we review the lessons learned from training population optimization in plants and the major challenges associated with the optimization of GS including population size, the relationship between training and test set (TS), update of TRS, and the use of different packages and algorithms for TRS implementation in GS. Finally, we describe general guidelines to improving the rate of genetic improvement by maximizing the use of the TRS optimization in the GS framework.