Search

The Aegilops tauschii-derived leaf rust resistance gene Lr42 has been widely used for breeding resistance wheat cultivars, but the molecular basis is unknown. Here, the authors show that Lr42 encodes an NLR-type of disease resistance gene by bulked segregant mapping in Ae. tauschii and confirm its function in common wheat.

Abstract The success of genomic selection (GS) in breeding schemes relies on its ability to provide accurate predictions of unobserved lines at early stages. Multigeneration data provides opportunities to increase the training data size and thus, the likelihood of extracting useful information from ancestors to improve prediction accuracy. The genomic best linear unbiased predictions (GBLUPs) are performed by borrowing information through kinship relationships between individuals. Multigeneration data usually becomes heterogeneous with complex family relationship patterns that are increasingly entangled with each generation. Under these conditions, historical data may not be optimal for model training as the accuracy could be compromised. The sparse selection index (SSI) is a method for training set (TRN) optimization, in which training individuals provide predictions to some but not all predicted subjects. We added an additional trimming process to the original SSI (trimmed SSI) to remove less important training individuals for prediction. Using a large multigeneration (8 yr) wheat (Triticum aestivum L.) grain yield dataset (n = 68,836), we found increases in accuracy as more years are included in the TRN, with improvements of ∼0.05 in the GBLUP accuracy when using 5 yr of historical data relative to when using only 1 yr. The SSI method showed a small gain over the GBLUP accuracy but with an important reduction on the TRN size. These reduced TRNs were formed with a similar number of subjects from each training generation. Our results suggest that the SSI provides a more stable ranking of genotypes than the GBLUP as the TRN becomes larger.

Spring bread wheat adaptation to diverse environments is supported by various traits such as phenology and plant architecture. A large-scale genome-wide association study (GWAS) was designed to investigate and dissect the genetic architecture of phenology affecting adaptation. It used 48 datasets from 4,680 spring wheat lines. For 8 years (2014–2021), these lines were evaluated for days to heading (DH) and maturity (DM) at three sites: Jabalpur, Ludhiana, and Samastipur (Pusa), which represent the three major Indian wheat-producing zones: the Central Zone (CZ), North-Western Plain Zone (NWPZ), and North-Eastern Plain Zone (NEPZ), respectively. Ludhiana had the highest mean DH of 103.8 days and DM of 148.6 days, whereas Jabalpur had the lowest mean DH of 77.7 days and DM of 121.6 days. We identified 119 markers significantly associated with DH and DM on chromosomes 5B (76), 2B (18), 7D (10), 4D (8), 5A (1), 6B (4), 7B (1), and 3D (1). Our results clearly indicated the importance of the photoperiod-associated gene (Ppd-B1) for adaptation to the NWPZ and the Vrn-B1 gene for adaptation to the NEPZ and CZ. A maximum variation of 21.1 and 14% was explained by markers 2B_56134146 and 5B_574145576 linked to the Ppd-B1 and Vrn-B1 genes, respectively, indicating their significant role in regulating DH and DM. The results provide important insights into the genomic regions associated with the two phenological traits that influence adaptation to the major wheat-producing zones in India.

Accelerating breeding efforts for developing biofortified bread wheat varieties necessitates understanding the genetic control of grain zinc concentration (GZnC) and grain iron concentration (GFeC). Hence, the major objective of this study was to perform genome-wide association mapping to identify consistently significant genotyping-by-sequencing markers associated with GZnC and GFeC using a large panel of 5,585 breeding lines from the International Maize and Wheat Improvement Center. These lines were grown between 2018 and 2021 in an optimally irrigated environment at Obregon, Mexico, while some of them were also grown in a water-limiting drought-stressed environment and a space-limiting small plot environment and evaluated for GZnC and GFeC. The lines showed a large and continuous variation for GZnC ranging from 27 to 74.5 ppm and GFeC ranging from 27 to 53.4 ppm. We performed 742,113 marker-traits association tests in 73 datasets and identified 141 markers consistently associated with GZnC and GFeC in three or more datasets, which were located on all wheat chromosomes except 3A and 7D. Among them, 29 markers were associated with both GZnC and GFeC, indicating a shared genetic basis for these micronutrients and the possibility of simultaneously improving both. In addition, several significant GZnC and GFeC associated markers were common across the irrigated, water-limiting drought-stressed, and space-limiting small plots environments, thereby indicating the feasibility of indirect selection for these micronutrients in either of these environments. Moreover, the many significant markers identified had minor effects on GZnC and GFeC, suggesting a quantitative genetic control of these traits. Our findings provide important insights into the complex genetic basis of GZnC and GFeC in bread wheat while implying limited prospects for marker-assisted selection and the need for using genomic selection.

Abstract Wheat grain yield (GY) improvement using genomic tools is important for achieving yield breakthroughs. To dissect the genetic architecture of wheat GY potential and stress-resilience, we have designed this large-scale genome-wide association study using 100 datasets, comprising 105,000 GY observations from 55,568 wheat lines evaluated between 2003 and 2019 by the International Maize and Wheat Improvement Center and national partners. We report 801 GY-associated genotyping-by-sequencing markers significant in more than one dataset and the highest number of them were on chromosomes 2A, 6B, 6A, 5B, 1B and 7B. We then used the linkage disequilibrium (LD) between the consistently significant markers to designate 214 GY-associated LD-blocks and observed that 84.5% of the 58 GY-associated LD-blocks in severe-drought, 100% of the 48 GY-associated LD-blocks in early-heat and 85.9% of the 71 GY-associated LD-blocks in late-heat, overlapped with the GY-associated LD-blocks in the irrigated-bed planting environment, substantiating that simultaneous improvement for GY potential and stress-resilience is feasible. Furthermore, we generated the GY-associated marker profiles and analyzed the GY favorable allele frequencies for a large panel of 73,142 wheat lines, resulting in 44.5 million datapoints. Overall, the extensive resources presented in this study provide great opportunities to accelerate breeding for high-yielding and stress-resilient wheat varieties.

Spot blotch caused by the fungus Bipolaris sorokiniana poses a serious threat to bread wheat production in warm and humid wheat-growing regions of the world. Hence, the major objective of this study was to identify consistent genotyping-by-sequencing (GBS) markers associated with spot blotch resistance using genome-wide association mapping on a large set of 6,736 advanced bread wheat breeding lines from the International Maize and Wheat Improvement Center. These lines were phenotyped as seven panels at Agua Fria, Mexico between the 2013–2014 and 2019–2020 crop cycles. We identified 214 significant spot blotch associated GBS markers in all the panels, among which only 96 were significant in more than one panel, indicating a strong environmental effect on the trait and highlights the need for multiple phenotypic evaluations to identify lines with stable spot blotch resistance. The 96 consistent GBS markers were on chromosomes 1A, 1B, 1D, 2A, 3B, 4A, 5B, 5D, 6B, 7A, 7B, and 7D, including markers possibly linked to the Lr46, Sb1, Sb2 and Sb3 genes. We also report the association of the 2NS translocation from Aegilops ventricosa with spot blotch resistance in some environments. Moreover, the spot blotch favorable alleles at the 2NS translocation and two markers on chromosome 3BS (3B_2280114 and 3B_5601689) were associated with increased grain yield evaluated at several environments in Mexico and India, implying that selection for favorable alleles at these loci could enable simultaneous improvement for high grain yield and spot blotch resistance. Furthermore, a significant relationship between the percentage of favorable alleles in the lines and their spot blotch response was observed, which taken together with the multiple minor effect loci identified to be associated with spot blotch in this study, indicate quantitative genetic control of resistance. Overall, the results presented here have extended our knowledge on the genetic basis of spot blotch resistance in bread wheat and further efforts to improve genetic resistance to the disease are needed for reducing current and future losses under climate change.

Since the inception of the theory and conceptual framework of genomic selection (GS), extensive research has been done on evaluating its efficiency for utilization in crop improvement. Though, the marker-assisted selection has proven its potential for improvement of qualitative traits controlled by one to few genes with large effects. Its role in improving quantitative traits controlled by several genes with small effects is limited. In this regard, GS that utilizes genomic-estimated breeding values of individuals obtained from genome-wide markers to choose candidates for the next breeding cycle is a powerful approach to improve quantitative traits. In the last two decades, GS has been widely adopted in animal breeding programs globally because of its potential to improve selection accuracy, minimize phenotyping, reduce cycle time, and increase genetic gains. In addition, given the promising initial evaluation outcomes of GS for the improvement of yield, biotic and abiotic stress tolerance, and quality in cereal crops like wheat, maize, and rice, prospects of integrating it in breeding crops are also being explored. Improved statistical models that leverage the genomic information to increase the prediction accuracies are critical for the effectiveness of GS-enabled breeding programs. Study on genetic architecture under drought and heat stress helps in developing production markers that can significantly accelerate the development of stress-resilient crop varieties through GS. This review focuses on the transition from traditional selection methods to GS, underlying statistical methods and tools used for this purpose, current status of GS studies in crop plants, and perspectives for its successful implementation in the development of climate-resilient crops.

Biofortification of cereal grains offers a lasting solution to combat micronutrient deficiency in developing countries where it poses developmental risks to children. Breeding efforts thus far have been directed toward increasing the grain concentrations of iron (Fe) and zinc (Zn) ions. Phytic acid (PA) chelates these metal ions, reducing their bioavailability in the digestive tract. We present a high-throughput assay for quantification of PA and its application in screening a breeding population. After extraction in 96-well megatiter plates, PA content was determined from the phosphate released after treatment with a commercially available phytase enzyme. In a set of 330 breeding lines of wheat grown in the field over 3 years as part of a HarvestPlus breeding program for high grain Fe and Zn, our assay unraveled variation for PA that ranged from 0.90 to 1.72% with a mean of 1.24%. PA content was not associated with grain yield. High yielding lines were further screened for low molar PA/Fe and PA/Zn ratios for increased metal ion bioavailability, demonstrating the utility of our assay. Genome-wide association study revealed 21 genetic associations, six of which were consistent across years. Five of these associations mapped to chromosomes 1A, 2A, 2D, 5A, and 7D. Additivity over four of these haplotypes accounted for an ∼10% reduction in PA. Our study demonstrates it is possible to scale up assays to directly select for low grain PA in forward breeding programs.

Wheat ranks first in the production and productivity of staple cereal crops in the world. Several diseases, including Stripe (Puccinia striiformis f. Sp. tritici), Black (Puccinia graminis f. Sp. tritici), and Brown (Puccinia recondita), have a major negative impact on wheat output, with 20 to 80% loss annually. Growing rust-resistant varieties is the most durable, cost-effective, and environmentally friendly way to combat rust pathogens. In the present review, we provide updated information on all black stem rust, yellow leaf rust, and brown leaf rust resistance genes including chromosomal position, those derived from different sources, nature of resistance type, and amino acid changes done by this gene against rust pathogen. This study summarized the 68 black stem rust, 101 leaf rust, and 108 stripe rust resistance genes from diverse cultivars of wheat and wheat primary and secondary gene pools. This review will be valuable to wheat breeders in cloning rust-resistant genes and developing leaf as well as stem rust-resistant wheat cultivars using gene pyramiding as well as frequency multiplication through introgression of the gene of interest for disease-free, sustainable grain production of wheat. The success of pyramiding genes from other sources to bread wheat depends on the nature of germplasm, the gap between flanking marker and targeted genes, the selection of genotypes in each generation, large number of gentoyes large genotype-environment interaction, etc., which is the future area of study.

Wheat blast is an emerging threat to wheat production, due to its recent migration to South Asia and Sub-Saharan Africa. Because genomic selection (GS) has emerged as a promising breeding strategy, the key objective of this study was to evaluate it for wheat blast phenotyped at precision phenotyping platforms in Quirusillas (Bolivia), Okinawa (Bolivia) and Jashore (Bangladesh) using three panels: (i) a diversity panel comprising 172 diverse spring wheat genotypes, (ii) a breeding panel comprising 248 elite breeding lines, and (iii) a full-sibs panel comprising 298 full-sibs. We evaluated two genomic prediction models (the genomic best linear unbiased prediction or GBLUP model and the Bayes B model) and compared the genomic prediction accuracies with accuracies from a fixed effects model (with selected blast-associated markers as fixed effects), a GBLUP + fixed effects model and a pedigree relationships-based model (ABLUP). On average, across all the panels and environments analyzed, the GBLUP + fixed effects model (0.63 ± 0.13) and the fixed effects model (0.62 ± 0.13) gave the highest prediction accuracies, followed by the Bayes B (0.59 ± 0.11), GBLUP (0.55 ± 0.1), and ABLUP (0.48 ± 0.06) models. The high prediction accuracies from the fixed effects model resulted from the markers tagging the 2NS translocation that had a large effect on blast in all the panels. This implies that in environments where the 2NS translocation-based blast resistance is effective, genotyping one to few markers tagging the translocation is sufficient to predict the blast response and genome-wide markers may not be needed. We also observed that marker-assisted selection (MAS) based on a few blast-associated markers outperformed GS as it selected the highest mean percentage (88.5%) of lines also selected by phenotypic selection and discarded the highest mean percentage of lines (91.8%) also discarded by phenotypic selection, across all panels. In conclusion, while this study demonstrates that MAS might be a powerful strategy to select for the 2NS translocation-based blast resistance, we emphasize that further efforts to use genomic tools to identify non-2NS translocation-based blast resistance are critical.

Tan spot (TS) and Septoria nodorum blotch (SNB), caused by Pyrenophora tritici-repentis and Parastagonospora nodorum, respectively, are important fungal leaf-spotting diseases of wheat that cause significant losses in grain yield. In this study, two recombinant inbred line populations, ‘Bartai’ × ‘Ciano T79’ (referred to as B × C) and ‘Cascabel’ × ‘Ciano T79’ (C × C) were tested for TS and SNB response in order to determine the genetic basis of seedling resistance. Genotyping was performed with the DArTseq genotyping-by-sequencing (GBS) platform. A chromosome region on 5AL conferred resistance to TS and SNB in both populations, but the effects were larger in B × C (R2 = 11.2%–16.8%) than in C × C (R2 = 2.5%–9.7%). Additionally, the chromosome region on 5BL (presumably Tsn1) was significant for both TS and SNB in B × C but not in C × C. Quantitative trait loci (QTL) with minor effects were identified on chromosomes 1B, 2A, 2B, 3A, 3B, 4D, 5A, 5B, 5D, 6B, and 6D. The two CIMMYT breeding lines ‘Bartai’ and ‘Cascabel’ contributed resistance alleles at both 5AL and 5BL QTL mentioned above. The QTL on 5AL showed linkage with the Vrn-A1 locus, whereas the vrn-A1 allele conferring lateness was associated with resistance to TS and SNB. Keywords: Parastagonospora nodorum, Pyrenophora tritici-repentis, QTL mapping, Resistance breeding, Triticum aestivum

Wheat blast (WB) caused by Magnaporthe oryzae pathotype Triticum (MoT) is an important fungal disease in tropical and subtropical wheat production regions. The disease was initially identified in Brazil in 1985, and it subsequently spread to some major wheat-producing areas of the country as well as several South American countries such as Bolivia, Paraguay, and Argentina. In recent years, WB has been introduced to Bangladesh and Zambia via international wheat trade, threatening wheat production in South Asia and Southern Africa with the possible further spreading in these two continents. Resistance source is mostly limited to 2NS carriers, which are being eroded by newly emerged MoT isolates, demonstrating an urgent need for identification and utilization of non-2NS resistance sources. Fungicides are also being heavily relied on to manage WB that resulted in increasing fungal resistance, which should be addressed by utilization of new fungicides or rotating different fungicides. Additionally, quarantine measures, cultural practices, non-fungicidal chemical treatment, disease forecasting, biocontrol etc., are also effective components of integrated WB management, which could be used in combination with varietal resistance and fungicides to obtain reasonable management of this disease.

Wheat (Triticum aestivum L.) is the most widely grown cereal crop in the world and is staple food to half the world’s population. The current world population is expected to reach 9.8 billion people by 2050, but food production is not expected to keep pace with demand in developing countries. Significant opportunities exist for traditional grain exporters to produce and export greater amounts of wheat to fill the gap. Karnal bunt, however, is a major threat, due to its use as a non-tariff trade barrier by several wheat-importing countries. The cultivation of resistant varieties remains the most cost-effective approach to manage the disease, but in countries that are free of the disease, genetic improvement is difficult due to quarantine restrictions. Here we report a study on pre-emptive breeding designed to identify linked molecular markers, evaluate the prospects of genomic selection as a tool, and prioritise wheat genotypes suitable for use as parents. In a genome-wide association (GWAS) study, we identified six DArTseq markers significantly linked to Karnal bunt resistance, which explained between 7.6 and 29.5% of the observed phenotypic variation. The accuracy of genomic prediction was estimated to vary between 0.53 and 0.56, depending on whether it is based solely on the identified Quantitative trait loci (QTL) markers or the use of genome-wide markers. As genotypes used as parents would be required to possess good yield and phenology, further research was conducted to assess the agronomic value of Karnal bunt resistant germplasm from the International Maize and Wheat Improvement Center (CIMMYT). We identified an ideal genotype, ZVS13_385, which possessed similar agronomic attributes to the highly successful Australian wheat variety, Mace. It is phenotypically resistant to Karnal bunt infection (

Wheat blast (WB) is a destructive disease in South America and its first outbreak in Bangladesh in 2016 posed a great risk to food security of South Asian countries. A genome wide association study (GWAS) was conducted on a diverse panel of 184 wheat genotypes from South Asia and CIMMYT. Phenotyping was conducted in eight field experiments in Bolivia and Bangladesh and a greenhouse experiment in the United States. Genotypic data included 11,401 SNP markers of the Illumina Infinium 15K BeadChip and four additional STS markers on the 2NS/2AS translocation region. Accessions with stable WB resistance across experiments were identified, which were all 2NS carriers. Nevertheless, a dozen moderately resistant 2AS lines were identified, exhibiting big variation among experiments. Significant marker-trait associations (MTA) were detected on chromosomes 1BS, 2AS, 6BS, and 7BL; but only MTAs on 2AS at the 2NS/2AS translocation region were consistently significant across experiments. The resistant accessions identified in this study could be used in production in South Asian countries as a preemptive strategy to prevent WB outbreak.

In this study, we defined the target population of environments (TPE) for wheat breeding in India, the largest wheat producer in South Asia, and estimated the correlated response to the selection and prediction ability of five selection environments (SEs) in Mexico. We also estimated grain yield (GY) gains in each TPE. Our analysis used meteorological, soil, and GY data from the international Elite Spring Wheat Yield Trials (ESWYT) distributed by the International Maize and Wheat Improvement Center (CIMMYT) from 2001 to 2016. We identified three TPEs: TPE 1, the optimally irrigated Northwestern Plain Zone; TPE 2, the optimally irrigated, heat-stressed North Eastern Plains Zone; and TPE 3, the drought-stressed Central-Peninsular Zone. The correlated response to selection ranged from 0.4 to 0.9 within each TPE. The highest prediction accuracies for GY per TPE were derived using models that included genotype-by-environment interaction and/or meteorological information and their interaction with the lines. The highest prediction accuracies for TPEs 1, 2, and 3 were 0.37, 0.46, and 0.51, respectively, and the respective GY gains were 118, 46, and 123 kg/ha/year. These results can help fine-tune the breeding of elite wheat germplasm with stable yields to reduce farmers’ risk from year-to-year environmental variation in India’s wheat lands, which cover 30 million ha, account for 100 million tons of grain or more each year, and provide food and livelihoods for hundreds of millions of farmers and consumers in South Asia.

Genome-enabled prediction plays an essential role in wheat breeding because it has the potential to increase the rate of genetic gain relative to traditional phenotypic and pedigree-based selection. Since the performance of wheat lines is highly influenced by environmental stimuli, it is important to accurately model the environment and its interaction with genetic factors in prediction models. Arguably, multi-environmental best linear unbiased prediction (BLUP) may deliver better prediction performance than single-environment genomic BLUP. We evaluated pedigree and genome-based prediction using 35,403 wheat lines from the Global Wheat Breeding Program of the International Maize and Wheat Improvement Center (CIMMYT). We implemented eight statistical models that included genome-wide molecular marker and pedigree information as prediction inputs in two different validation schemes. All models included main effects, but some considered interactions between the different types of pedigree and genomic covariates via Hadamard products of similarity kernels. Pedigree models always gave better prediction of new lines in observed environments than genome-based models when only main effects were fitted. However, for all traits, the highest predictive abilities were obtained when interactions between pedigree, genomes, and environments were included. When new lines were predicted in unobserved environments, in almost all trait/year combinations, the marker main-effects model was the best. These results provide strong evidence that the different sources of genetic information (molecular markers and pedigree) are not equally useful at different stages of the breeding pipelines, and can be employed differentially to improve the design and prediction of the outcome of future breeding programs.

Kernel methods are flexible and easy to interpret and have been successfully used in genomic-enabled prediction of various plant species. Kernel methods used in genomic prediction comprise the linear genomic best linear unbiased predictor (GBLUP or GB) kernel, and the Gaussian kernel (GK). In general, these kernels have been used with two statistical models: single-environment and genomic × environment (GE) models. Recently near infrared spectroscopy (NIR) has been used as an inexpensive and non-destructive high-throughput phenotyping method for predicting unobserved line performance in plant breeding trials. In this study, we used a non-linear arc-cosine kernel (AK) that emulates deep learning artificial neural networks. We compared AK prediction accuracy with the prediction accuracy of GB and GK kernel methods in four genomic data sets, one of which also includes pedigree and NIR information. Results show that for all four data sets, AK and GK kernels achieved higher prediction accuracy than the linear GB kernel for the single-environment and GE multi-environment models. In addition, AK achieved similar or slightly higher prediction accuracy than the GK kernel. For all data sets, the GE model achieved higher prediction accuracy than the single-environment model. For the data set that includes pedigree, markers and NIR, results show that the NIR wavelength alone achieved lower prediction accuracy than the genomic information alone; however, the pedigree plus NIR information achieved only slightly lower prediction accuracy than the marker plus the NIR high-throughput data.

Multiple-trait experiments with mixed phenotypes (binary, ordinal and continuous) are not rare in animal and plant breeding programs. However, there is a lack of statistical models that can exploit the correlation between traits with mixed phenotypes in order to improve prediction accuracy in the context of genomic selection (GS). For this reason, when breeders have mixed phenotypes, they usually analyze them using univariate models, and thus are not able to exploit the correlation between traits, which many times helps improve prediction accuracy. In this paper we propose applying deep learning for analyzing multiple traits with mixed phenotype data in terms of prediction accuracy. The prediction performance of multiple-trait deep learning with mixed phenotypes (MTDLMP) models was compared to the performance of univariate deep learning (UDL) models. Both models were evaluated using predictors with and without the genotype × environment (G×E) interaction term (I and WI, respectively). The metric used for evaluating prediction accuracy was Pearson’s correlation for continuous traits and the percentage of cases correctly classified (PCCC) for binary and ordinal traits. We found that a modest gain in prediction accuracy was obtained only in the continuous trait under the MTDLMP model compared to the UDL model, whereas for the other traits (1 binary and 2 ordinal) we did not find any difference between the two models. In both models we observed that the prediction performance was better for WI than for I. The MTDLMP model is a good alternative for performing simultaneous predictions of mixed phenotypes (binary, ordinal and continuous) in the context of GS.

Genomic selection is revolutionizing plant breeding. However, still lacking are better statistical models for ordinal phenotypes to improve the accuracy of the selection of candidate genotypes. For this reason, in this paper we explore the genomic based prediction performance of two popular machine learning methods: the Multi Layer Perceptron (MLP) and support vector machine (SVM) methods vs. the Bayesian threshold genomic best linear unbiased prediction (TGBLUP) model. We used the percentage of cases correctly classified (PCCC) as a metric to measure the prediction performance, and seven real data sets to evaluate the prediction accuracy, and found that the best predictions (in four out of the seven data sets) in terms of PCCC occurred under the TGLBUP model, while the worst occurred under the SVM method. Also, in general we found no statistical differences between using 1, 2 and 3 layers under the MLP models, which means that many times the conventional neuronal network model with only one layer is enough. However, although even that the TGBLUP model was better, we found that the predictions of MLP and SVM were very competitive with the advantage that the SVM was the most efficient in terms of the computational time required.

We untangled key regions of the genetic architecture of grain yield (GY) in CIMMYT spring bread wheat by conducting a haplotype-based, genome-wide association study (GWAS), together with an investigation of epistatic interactions using seven large sets of elite yield trials (EYTs) consisting of a total of 6,461 advanced breeding lines. These lines were phenotyped under irrigated and stress environments in seven growing seasons (2011–2018) and genotyped with genotyping-by-sequencing markers. Genome-wide 519 haplotype blocks were constructed, using a linkage disequilibrium-based approach covering 14,036 Mb in the wheat genome. Haplotype-based GWAS identified 7, 4, 10, and 15 stable (significant in three or more EYTs) associations in irrigated (I), mild drought (MD), severe drought (SD), and heat stress (HS) testing environments, respectively. Considering all EYTs and the four testing environments together, 30 stable associations were deciphered with seven hotspots identified on chromosomes 1A, 1B, 2B, 4A, 5B, 6B, and 7B, where multiple haplotype blocks were associated with GY. Epistatic interactions contributed significantly to the genetic architecture of GY, explaining variation of 3.5–21.1%, 3.7–14.7%, 3.5–20.6%, and 4.4– 23.1% in I, MD, SD, and HS environments, respectively. Our results revealed the intricate genetic architecture of GY, controlled by both main and epistatic effects. The importance of these results for practical applications in the CIMMYT breeding program is discussed.

Abstract Wheat quality improvement is an important objective in all wheat breeding programs. However, due to the cost, time and quantity of seed required, wheat quality is typically analyzed only in the last stages of the breeding cycle on a limited number of samples. The use of genomic prediction could greatly help to select for wheat quality more efficiently by reducing the cost and time required for this analysis. Here were evaluated the prediction performances of 13 wheat quality traits under two multi‐trait models (Bayesian multi‐trait multi‐environment [BMTME] and multi‐trait ridge regression [MTR]) using five data sets of wheat lines evaluated in the field during two consecutive years. Lines in the second year (testing) were predicted using the quality information obtained in the first year (training). For most quality traits were found moderate to high prediction accuracies, suggesting that the use of genomic selection could be feasible. The best predictions were obtained with the BMTME model in all traits and the worst with the MTR model. The best predictions with the BMTME model under the mean arctangent absolute percentage error (MAAPE) were for test weight across the five data sets, whereas the worst predictions were for the alveograph trait ALVPL. In contrast, under Pearson's correlation, the best predictions depended on the data set. The results obtained suggest that the BMTME model should be preferred for multi‐trait prediction analyses. This model allows to obtain not only the correlation among traits, but also the correlation among environments, helping to increase the prediction accuracy.

Wheat blast (WB) is a devastating fungal disease that has recently spread to Bangladesh and poses a threat to the wheat production in India, which is the second-largest wheat producing country in the world. In this study, 350 Indian wheat genotypes were evaluated for WB resistance in 12 field experiments in three different locations, namely Jashore in Bangladesh and Quirusillas and Okinawa in Bolivia. Single nucleotide polymorphisms (SNPs) across the genome were obtained using DArTseq® technology, and 7554 filtered SNP markers were selected for a genome-wide association study (GWAS). All the three GWAS approaches used identified the 2NS translocation as the only major source of resistance, explaining up to 32% of the phenotypic variation. Additional marker-trait associations were located on chromosomes 2B, 3B, 4D, 5A and 7A, and the combined effect of three SNPs (2B_180938790, 7A_752501634 and 5A_618682953) showed better resistance, indicating their additive effects on WB resistance. Among the 298 bread wheat genotypes, 89 (29.9%) carried the 2NS translocation, the majority of which (60 genotypes) were CIMMYT introductions, and 29 were from India. The 2NS carriers with a grand mean WB index of 6.6 showed higher blast resistance compared to the non-2NS genotypes with a mean index of 46.5. Of the 52 durum wheats, only one genotype, HI 8819, had the 2NS translocation and was the most resistant, with a grand mean WB index of 0.93. Our study suggests that the 2NS translocation is the only major resistance source in the Indian wheat panel analysed and emphasizes the urgent need to identify novel non-2NS resistance sources and genomic regions.

Genomic breeding technologies offer new opportunities for grain yield (GY) improvement in common wheat. In this study, we have evaluated the potential of genomic selection (GS) in breeding for GY in wheat by modeling a large dataset of 48,562 GY observations from the International Maize and Wheat Improvement Center (CIMMYT), including 36 yield trials evaluated between 2012 and 2019 in Obregón, Sonora, Mexico. Our key objective was to determine the value that GS can add to the current three-stage yield testing strategy at CIMMYT, and we draw inferences from predictive modeling of GY using 420 different populations, environments, cycles, and model combinations. First, we evaluated the potential of genomic predictions for minimizing the number of replications and lines tested within a site and year and obtained mean prediction accuracies (PAs) of 0.56, 0.5, and 0.42 in Stages 1, 2, and 3 of yield testing, respectively. However, these PAs were similar to the mean pedigree-based PAs indicating that genomic relationships added no value to pedigree relationships in the yield testing stages, characterized by small family-sizes. Second, we evaluated genomic predictions for minimizing GY testing across stages/years in Obregón and observed mean PAs of 0.41, 0.31, and 0.37, respectively when GY in the full irrigation bed planting (FI BP), drought stress (DS), and late-sown heat stress environments were predicted across years using genotype × environment (G × E) interaction models. Third, we evaluated genomic predictions for minimizing the number of yield testing environments and observed that in Stage 2, the FI BP, full irrigation flat planting and early-sown heat stress environments (mean PA of 0.37 ± 0.12) and the reduced irrigation and DS environments (mean PA of 0.45 ± 0.07) had moderate predictabilities among them. However, in both predictions across years and environments, the PAs were inconsistent across years and the G × E models had no advantage over the baseline model with environment and line effects. Overall, our results provide excellent insights into the predictability of a quantitative trait like GY and will have important implications on the future design of GS for GY in wheat breeding programs globally.

Tan spot caused by Pyrenophora tritici-repentis (Ptr) is an important disease of wheat in many wheat producing areas of the world. A genome wide association study (GWAS) was conducted using 11,401 SNP markers of the Illumina Infinium 15K Bead Chip with whole genome coverage to identify genomic regions associated with resistance to tan spot in a diverse panel of 184 wheat genotypes originating from South Asia and CIMMYT. The GWAS panel was phenotyped for seedling resistance to tan spot with Ptr race 1 in two greenhouse experiments. Besides CIMMYT germplasm, several lines from South Asia (India, Bangladesh and Nepal) showed good degree of resistance to tan spot. Association mapping was conducted separately for individual experiments and for pooled data using mixed linear model (MLM) and Fixed and random model Circulating Probability Unification (FarmCPU) model; no significant MTAs were recorded through the MLM model, whereas FarmCPU model reported nine significant MTAs located on chromosomes 1B, 2A, 2B, 3B, 4A, 5A, 5B, 6A, and 7D. The long arms of chromosomes 5A and 5B were consistent across both environments, in which the Vrn-A1 locus was found in identified region of chromosome 5A, and MTA at IACX9261 on 5BL appears to represent the resistance gene tsn 1. MTAs observed on chromosomes 1B, 2A, 2B, 3B, 4A, 6A, and 7D have not been reported previously and are likely novel.

Breeding for grain yield (GY) in bread wheat at the International Maize and Wheat Improvement Center (CIMMYT) involves three-stage testing at Obregon, Mexico in different selection environments (SEs). To understand the efficiency of selection in the SEs, we performed a large retrospective quantitative genetics study using CIMMYT’s yield trials evaluated in the SEs (2013–2014 to 2017–2018), the South Asia Bread Wheat Genomic Prediction Yield Trials (SABWGPYTs) evaluated in India, Pakistan, and Bangladesh (2014–2015 to 2017–2018), and the Elite Spring Wheat Yield Trials (ESWYTs) evaluated in several sites globally (2003–2004 to 2016–2017). First, we compared the narrow-sense heritabilities in the Obregon SEs and target sites and observed that the mean heritability in the SEs was 44.2 and 92.3% higher than the mean heritabilities in the SABWGPYT and ESWYT sites, respectively. Second, we observed significant genetic correlations between a SE in Obregon and all the five SABWGPYT sites and 65.1% of the ESWYT sites. Third, we observed high ratios of response to indirect selection in the SEs of Obregon with a mean of 0.80 ± 0.21 and 2.6 ± 5.4 in the SABWGPYT and ESWYT sites, respectively. Furthermore, our results also indicated that for all the SABWGPYT sites and 82% of the ESWYT sites, a response greater than 0.5 can be achieved by indirect selection for GY in Obregon. We also performed genomic prediction for GY in the target sites using the performance of the same lines in the SEs of Obregon and observed moderate mean prediction accuracies of 0.24 ± 0.08 and 0.28 ± 0.08 in the SABWGPYT and ESWYT sites, respectively using the genotype x environment (GxE) model. However, we observed similar accuracies using the baseline model with environment and line effects and no advantage of modeling GxE interactions. Overall, this study provides important insights into the suitability of the Obregon SEs in breeding for GY, while the variable genomic predictabilities of GY and the high year-to-year GY fluctuations reported, highlight the importance of multi-environment testing across time and space to stave off GxE induced uncertainties in varietal yields.

Deep learning (DL) is a promising method for genomic-enabled prediction. However, the implementation of DL is difficult because many hyperparameters (number of hidden layers, number of neurons, learning rate, number of epochs, batch size, etc.) need to be tuned. For this reason, deep kernel methods, which only require defining the number of layers, may be an attractive alternative. Deep kernel methods emulate DL models with a large number of neurons, but are defined by relatively easily computed covariance matrices. In this research, we compared the genome-based prediction of DL to a deep kernel (arc-cosine kernel, AK), to the commonly used non-additive Gaussian kernel (GK), as well as to the conventional additive genomic best linear unbiased predictor (GBLUP/GB). We used two real wheat data sets for benchmarking these methods. On average, AK and GK outperformed DL and GB. The gain in terms of prediction performance of AK and GK over DL and GB was not large, but AK and GK have the advantage that only one parameter, the number of layers (AK) or the bandwidth parameter (GK), has to be tuned in each method. Furthermore, although AK and GK had similar performance, deep kernel AK is easier to implement than GK, since the parameter “number of layers” is more easily determined than the bandwidth parameter of GK. Comparing AK and DL for the data set of year 2015–2016, the difference in performance of the two methods was bigger, with AK predicting much better than DL. On this data, the optimization of the hyperparameters for DL was difficult and the finally used parameters may have been suboptimal. Our results suggest that AK is a good alternative to DL with the advantage that practically no tuning process is required.

In genomic-enabled prediction, the task of improving the accuracy of the prediction of lines in environments is difficult because the available information is generally sparse and usually has low correlations between traits. In current genomic selection, although researchers have a large amount of information and appropriate statistical models to process it, there is still limited computing efficiency to do so. Although some statistical models are usually mathematically elegant, many of them are also computationally inefficient, and they are impractical for many traits, lines, environments, and years because they need to sample from huge normal multivariate distributions. For these reasons, this study explores two recommender systems: item-based collaborative filtering (IBCF) and the matrix factorization algorithm (MF) in the context of multiple traits and multiple environments. The IBCF and MF methods were compared with two conventional methods on simulated and real data. Results of the simulated and real data sets show that the IBCF technique was slightly better in terms of prediction accuracy than the two conventional methods and the MF method when the correlation was moderately high. The IBCF technique is very attractive because it produces good predictions when there is high correlation between items (environment–trait combinations) and its implementation is computationally feasible, which can be useful for plant breeders who deal with very large data sets.

Wheat blast (WB) disease, since its first identification in Bangladesh in 2016, is now an established serious threat to wheat production in South Asia. There is a need for sound knowledge about resistance sources and associated genomic regions to assist breeding programs. Hence, a panel of genotypes from India and Bangladesh was evaluated for wheat blast resistance and a genome-wide association study (GWAS) was performed. Disease evaluation was done during five crop seasons—at precision phenotyping platform (PPPs) for wheat blast disease at Jashore (2018–19), Quirusillas (2018–19 and 2019–20) and Okinawa (2019 and 2020). Single nucleotide polymorphisms (SNP) across the genome were obtained using DArTseq genotyping-by-sequencing platform, and in total 5713 filtered markers were used. GWAS revealed 40 significant markers associated with WB resistance, of which 33 (82.5%) were in the 2NS/2AS chromosome segment and one each on seven chromosomes (3B, 3D, 4A, 5A, 5D, 6A and 6B). The 2NS markers contributed significantly in most of the environments, explaining an average of 33.4% of the phenotypic variation. Overall, 22.4% of the germplasm carried 2NS/2AS segment. So far, 2NS translocation is the only effective WB resistance source being used in the breeding programs of South Asia. Nevertheless, the identification of non-2NS/2AS genomic regions for WB resistance provides a hope to broaden and diversify resistance for this disease in years to come.

When a plant scientist wishes to make genomic-enabled predictions of multiple traits measured in multiple individuals in multiple environments, the most common strategy for performing the analysis is to use a single trait at a time taking into account genotype × environment interaction (G × E), because there is a lack of comprehensive models that simultaneously take into account the correlated counting traits and G × E. For this reason, in this study we propose a multiple-trait and multiple-environment model for count data. The proposed model was developed under the Bayesian paradigm for which we developed a Markov Chain Monte Carlo (MCMC) with noninformative priors. This allows obtaining all required full conditional distributions of the parameters leading to an exact Gibbs sampler for the posterior distribution. Our model was tested with simulated data and a real data set. Results show that the proposed multi-trait, multi-environment model is an attractive alternative for modeling multiple count traits measured in multiple environments.

Genomic selection (GS) has been promising for increasing genetic gains in several species. Therefore, we evaluated the potential integration of GS for grain yield (GY) in bread wheat ( L.) in CIMMYT’s elite yield trial nurseries. We observed that the genomic prediction accuracies within nurseries (0.44 and 0.35) were substantially higher than across-nursery accuracies (0.15 and 0.05) for GY evaluated in the bed and flat planting systems, respectively. The accuracies from using only a subset of 251 genotyping-by-sequencing markers were comparable to the accuracies using all 2038 markers. We also used the item-based collaborative filtering approach for incorporating other related traits in predicting GY and observed that it outperformed genomic predictions across nurseries, but was less predictive when trait correlations with GY were low. Furthermore, we compared GS and phenotypic selections (PS) and observed that at a selection intensity of 0.5, GS could select a maximum of 70.9 and 61.5% of the top lines and discard 71.5 and 60.5% of the poor lines selected or discarded by PS within and across nurseries, respectively. Comparisons of GS and pedigree-based predictions revealed that the advantage of GS over the pedigree was moderate in populations without full-sibs. However, GS was less advantageous for within-family selections in elite families with few full-sibs and minimal Mendelian sampling variance. Overall, our results demonstrate the importance of applying GS for GY at the appropriate stage of the breeding cycle, and we speculate that gains can be maximized if it is implemented in early-generation within-family selections.

The Item-Based Collaborative Filtering for Multitrait and Multienvironment Data (IBCF.MTME) package was developed to implement the item-based collaborative filtering (IBCF) algorithm for continuous phenotypic data in the context of plant breeding where data are collected for various traits and environments. The main difference between this package and the other available packages that can implement IBCF is that this one was developed for continuous phenotypic data, which cannot be implemented in the current packages because they can implement IBCF only for binary and ordinary phenotypes. In the following article, we will show how to both install the package and use it for studying the prediction accuracy of multitrait and multienvironment data under phenotypic and genomic selection. We illustrate its use with seven examples (with information from two datasets, Wheat_IBCF and Year_IBCF, which are included in the package) comprising multienvironment data, multitrait data, and both multitrait and multienvironment data that cover scenarios in which breeding scientists are interested. The package offers many advantages for studying the genomic-enabled prediction accuracy of multitrait and multienvironment data, ultimately helping plant breeders make better decisions.

The leaf spotting diseases in wheat that include Septoria tritici blotch (STB) caused by , Stagonospora nodorum blotch (SNB) caused by , and tan spot (TS) caused by pose challenges to breeding programs in selecting for resistance. A promising approach that could enable selection prior to phenotyping is genomic selection that uses genome-wide markers to estimate breeding values (BVs) for quantitative traits. To evaluate this approach for seedling and/or adult plant resistance (APR) to STB, SNB, and TS, we compared the predictive ability of least-squares (LS) approach with genomic-enabled prediction models including genomic best linear unbiased predictor (GBLUP), Bayesian ridge regression (BRR), Bayes A (BA), Bayes B (BB), Bayes Cπ (BC), Bayesian least absolute shrinkage and selection operator (BL), and reproducing kernel Hilbert spaces markers (RKHS-M), a pedigree-based model (RKHS-P) and RKHS markers and pedigree (RKHS-MP). We observed that LS gave the lowest prediction accuracies and RKHS-MP, the highest. The genomic-enabled prediction models and RKHS-P gave similar accuracies. The increase in accuracy using genomic prediction models over LS was 48%. The mean genomic prediction accuracies were 0.45 for STB (APR), 0.55 for SNB (seedling), 0.66 for TS (seedling) and 0.48 for TS (APR). We also compared markers from two whole-genome profiling approaches: genotyping by sequencing (GBS) and diversity arrays technology sequencing (DArTseq) for prediction. While, GBS markers performed slightly better than DArTseq, combining markers from the two approaches did not improve accuracies. We conclude that implementing GS in breeding for these diseases would help to achieve higher accuracies and rapid gains from selection.

Karnal bunt disease of wheat, caused by the fungus Neovossia indica, is one of the most important challenges to the grain industry as it affects the grain quality and also restricts the international movement of infected grain. It is a seed-, soil- and airborne disease with limited effect of chemical control. Currently, this disease is contained through the deployment of host resistance but further improvement is limited as only a few genotypes have been found to carry partial resistance. To identify genomic regions responsible for resistance in a set of 339 wheat accessions, genome-wide association study (GWAS) was undertaken using the DArTSeq® technology, in which 18 genomic regions for Karnal bunt resistance were identified, explaining 5−20% of the phenotypic variation. The identified quantitative trait loci (QTL) on chromosome 2BL showed consistently significant effects across all four experiments, whereas another QTL on 5BL was significant in three experiments. Additional QTLs were mapped on chromosomes 1DL, 2DL, 4AL, 5AS, 6BL, 6BS, 7BS and 7DL that have not been mapped previously, and on chromosomes 4B, 5AL, 5BL and 6BS, which have been reported in previous studies. Germplasm with less than 1% Karnal bunt infection have been identified and can be used for resistance breeding. The SNP markers linked to the genomic regions conferring resistance to Karnal bunt could be used to improve Karnal bunt resistance through marker-assisted selection.

Spot blotch (SB) is an important fungal disease of wheat in South Asia and South America. Host resistance is regarded as an economical and environmentally friendly approach of controlling SB, and the inheritance of resistance is mostly quantitative. In order to gain a better understanding on the SB resistance mechanism in CIMMYT germplasm, two bi-parental mapping populations were generated, both comprising 232 F2:7 progenies. Elite CIMMYT breeding lines, BARTAI and WUYA, were used as resistant parents, whereas CIANO T79 was used as susceptible parent in both populations. The two populations were evaluated for field SB resistance at CIMMYT’s Agua Fria station for three consecutive years, from the 2012⁻2013 to 2014⁻2015 cropping seasons. Phenological traits like plant height (PH) and days to heading (DH) were also determined. Genotyping was performed using the DArTSeq genotyping-by-sequencing (GBS) platform, and a few D-genome specific SNPs and those for phenological traits were integrated for analysis. The most prominent quantitative trait locus (QTL) in both populations was found on chromosome 5AL at the Vrn-A1 locus, explaining phenotypic variations of 7⁻27%. Minor QTL were found on chromosomes 1B, 3A, 3B, 4B, 4D, 5B and 6D in BARTAI and on chromosomes 1B, 2A, 2D and 4B in WUYA, whereas minor QTL contributed by CIANO T79 were identified on chromosome 1B, 1D, 3A, 4B and 7A. In summary, resistance to SB in the two mapping populations was controlled by multiple minor QTL, with strong influence from Vrn-A1.