Metabolomic prediction of yield in hybrid rice

Molecular concepts to explain heterosis in crops

Heterosis describes the superior performance of hybrid plants compared with their genetically distinct parents and is a pillar of global food security. Here we review the current status of the molecular dissection of heterosis. We discuss how extensive intraspecific structural genomic variation between parental genotypes leads to heterosis by genetic complementation in hybrids. Moreover, we survey how global gene expression complementation contributes to heterosis by hundreds of additionally active genes in hybrids and how overdominant single genes mediate heterosis in several species. Furthermore, we highlight the prominent role of the microbiome in improving the performance of hybrids. Taken together, the molecular understanding of heterosis will pave the way to accelerate hybrid productivity and a more sustainable agriculture.

GA-GBLUP: leveraging the genetic algorithm to improve the predictability of genomic selection

Genomic selection (GS) has emerged as an effective technology to accelerate crop hybrid breeding by enabling early selection prior to phenotype collection. Genomic best linear unbiased prediction (GBLUP) is a robust method that has been routinely used in GS breeding programs. However, GBLUP assumes that markers contribute equally to the total genetic variance, which may not be the case. In this study, we developed a novel GS method called GA-GBLUP that leverages the genetic algorithm (GA) to select markers related to the target trait. We defined four fitness functions for optimization, including AIC, BIC, R2, and HAT, to improve the predictability and bin adjacent markers based on the principle of linkage disequilibrium to reduce model dimension. The results demonstrate that the GA-GBLUP model, equipped with R2 and HAT fitness function, produces much higher predictability than GBLUP for most traits in rice and maize datasets, particularly for traits with low heritability. Moreover, we have developed a user-friendly R package, GAGBLUP, for GS, and the package is freely available on CRAN (https://CRAN.R-project.org/package=GAGBLUP).

Advances in the evolution research and genetic breeding of peanut

Peanut is an important cash crop used in oil, food and feed in our country. The rapid development of sequencing technology has promoted the research on the related aspects of peanut genetic breeding. This paper reviews the research progress of peanut origin and evolution, genetic breeding, molecular markers and their applications, genomics, QTL mapping and genome selection techniques. The main problems of molecular genetic breeding in peanut research worldwide include: the narrow genetic resources of cultivated species, unstable genetic transformation and unclear molecular mechanism of important agronomic traits. Considering the severe challenges regarding the supply of edible oil, and the main problems in peanut production, the urgent research directions of peanut are put forward: The de novo domestication and the exploitation of excellent genes from wild resources to improve modern cultivars; Integration of multi-omics data to enhance the importance of big data in peanut genetics and breeding; Cloning the important genes related to peanut agronomic traits and analyzing their fine regulation mechanisms; Precision molecular design breeding and using gene editing technology to accurately improve the key traits of peanut.

Multi-omics assists genomic prediction of maize yield with machine learning approaches

With the improvement of high-throughput technologies in recent years, large multi-dimensional plant omics data have been produced, and big-data-driven yield prediction research has received increasing attention. Machine learning offers promising computational and analytical solutions to interpret the biological meaning of large amounts of data in crops. In this study, we utilized multi-omics datasets from 156 maize recombinant inbred lines, containing 2496 single nucleotide polymorphisms (SNPs), 46 image traits (i-traits) from 16 developmental stages obtained through an automatic phenotyping platform, and 133 primary metabolites. Based on benchmark tests with different types of prediction models, some machine learning methods, such as Partial Least Squares (PLS), Random Forest (RF), and Gaussian process with Radial basis function kernel (GaussprRadial), achieved better prediction for maize yield, albeit slight difference for method preferences among i-traits, genomic, and metabolic data. We found that better yield prediction may be caused by various capabilities in ranking and filtering data features, which is found to be linked with biological meaning such as photosynthesis-related or kernel development-related regulations. Finally, by integrating multiple omics data with the RF machine learning approach, we can further improve the prediction accuracy of grain yield from 0.32 to 0.43. Our research provides new ideas for the application of plant omics data and artificial intelligence approaches to facilitate crop genetic improvements.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11032-024-01454-z.

Importance of genetic architecture in marker selection decisions for genomic prediction

KEY MESSAGE

We demonstrate potential for improved multi-environment genomic prediction accuracy using structural variant markers. However, the degree of observed improvement is highly dependent on the genetic architecture of the trait. Breeders commonly use genetic markers to predict the performance of untested individuals as a way to improve the efficiency of breeding programs. These genomic prediction models have almost exclusively used single nucleotide polymorphisms (SNPs) as their source of genetic information, even though other types of markers exist, such as structural variants (SVs). Given that SVs are associated with environmental adaptation and not all of them are in linkage disequilibrium to SNPs, SVs have the potential to bring additional information to multi-environment prediction models that are not captured by SNPs alone. Here, we evaluated different marker types (SNPs and/or SVs) on prediction accuracy across a range of genetic architectures for simulated traits across multiple environments. Our results show that SVs can improve prediction accuracy, but it is highly dependent on the genetic architecture of the trait and the relative gain in accuracy is minimal. When SVs are the only causative variant type, 70% of the time SV predictors outperform SNP predictors. However, the improvement in accuracy in these instances is only 1.5% on average. Further simulations with predictors in varying degrees of LD with causative variants of different types (e.g., SNPs, SVs, SNPs and SVs) showed that prediction accuracy increased as linkage disequilibrium between causative variants and predictors increased regardless of the marker type. This study demonstrates that knowing the genetic architecture of a trait in deciding what markers to use in large-scale genomic prediction modeling in a breeding program is more important than what types of markers to use.

A Bayesian model for genomic prediction using metabolic networks

Motivation

Genomic prediction is now an essential technique in breeding and medicine, and it is interesting to see how omics data can be used to improve prediction accuracy. Precedent work proposed a metabolic network-based method in biomass prediction of Arabidopsis; however, the method consists of multiple steps that possibly degrade prediction accuracy.

Results

We proposed a Bayesian model that integrates all steps and jointly infers all fluxes of reactions related to biomass production. The proposed model showed higher accuracies than methods compared both in simulated and real data. The findings support the previous excellent idea that metabolic network information can be used for prediction.

Availability and implementation

All R and stan scripts to reproduce the results of this study are available at https://github.com/Onogi/MetabolicModeling.

QTL analysis of important agronomic traits and metabolites in foxtail millet (Setaria italica) by RIL population and widely targeted metabolome

As a bridge between genome and phenotype, metabolome is closely related to plant growth and development. However, the research on the combination of genome, metabolome and multiple agronomic traits in foxtail millet (Setaria italica) is insufficient. Here, based on the linkage analysis of 3,452 metabolites via with high-quality genetic linkage maps, we detected a total of 1,049 metabolic quantitative trait loci (mQTLs) distributed in 11 hotspots, and 28 metabolite-related candidate genes were mined from 14 mQTLs. In addition, 136 single-environment phenotypic QTL (pQTLs) related to 63 phenotypes were identified by linkage analysis, and there were 12 hotspots on these pQTLs. We futher dissected 39 candidate genes related to agronomic traits through metabolite-phenotype correlation and gene function analysis, including Sd1 semidwarf gene, which can affect plant height by regulating GA synthesis. Combined correlation network and QTL analysis, we found that flavonoid-lignin pathway maybe closely related to plant architecture and yield in foxtail millet. For example, the correlation coefficient between apigenin 7-rutinoside and stem diameter reached 0.98, and they were co-located at 41.33-44.15 Mb of chromosome 5, further gene function analysis revealed that 5 flavonoid pathway genes, as well as Sd1, were located in this interval . Therefore, the correlation and co-localization between flavonoid-lignins and plant architecture may be due to the close linkage of their regulatory genes in millet. Besides, we also found that a combination of genomic and metabolomic for BLUP analysis can better predict plant agronomic traits than genomic or metabolomic data, independently. In conclusion, the combined analysis of mQTL and pQTL in millet have linked genetic, metabolic and agronomic traits, and is of great significance for metabolite-related molecular assisted breeding.

Metabolite Profiling of Wheat Response to Cultivar Improvement and Nitrogen Fertilizer

Both genetic improvement and the application of N fertilizer increase the quality and yields of wheat. However, the molecular kinetics that underlies the differences between them are not well understood. In this study, we performed a non-targeted metabolomic analysis on wheat cultivars from different release years to comprehensively investigate the metabolic differences between cultivar and N treatments. The results revealed that the plant height and tiller number steadily decreased with increased ears numbers, whereas the grain number and weight increased with genetic improvement. Following the addition of N fertilizer, the panicle numbers and grain weights increased in an old cultivar, whereas the panicle number and grain number per panicle increased in a modern cultivar. For the 1950s to 2010s cultivar, the yield increases due to genetic improvements ranged from -1.9% to 96.7%, whereas that of N application ranged from 19.1% to 81.6%. Based on the untargeted metabolomics approach, the findings demonstrated that genetic improvements induced 1.4 to 7.4 times more metabolic alterations than N fertilizer supply. After the addition of N, 69.6%, 29.4%, and 33.3% of the differential metabolites were upregulated in the 1950s, 1980s, and 2010s cultivars, respectively. The results of metabolic pathway analysis of the identified differential metabolites via genetic improvement indicated enrichment in 1-2 KEGG pathways, whereas the application of N fertilizer enriched 2-4 pathways. Our results provide new insights into the molecular mechanisms of wheat quality and grain yield developments.

Multi-omics revolution to promote plant breeding efficiency

Crop production is the primary goal of agricultural activities, which is always taken into consideration. However, global agricultural systems are coming under increasing pressure from the rising food demand of the rapidly growing world population and changing climate. To address these issues, improving high-yield and climate-resilient related-traits in crop breeding is an effective strategy. In recent years, advances in omics techniques, including genomics, transcriptomics, proteomics, and metabolomics, paved the way for accelerating plant/crop breeding to cope with the changing climate and enhance food production. Optimized omics and phenotypic plasticity platform integration, exploited by evolving machine learning algorithms will aid in the development of biological interpretations for complex crop traits. The precise and progressive assembly of desire alleles using precise genome editing approaches and enhanced breeding strategies would enable future crops to excel in combating the changing climates. Furthermore, plant breeding and genetic engineering ensures an exclusive approach to developing nutrient sufficient and climate-resilient crops, the productivity of which can sustainably and adequately meet the world's food, nutrition, and energy needs. This review provides an overview of how the integration of omics approaches could be exploited to select crop varieties with desired traits.

Genome-wide identification of genes enabling accurate prediction of hybrid performance from parents across environments and populations for gene-based breeding in maize

Accurate prediction of hybrid offspring complex trait phenotype from parents is paramount to enhanced plant breeding, animal breeding, and human medicine. Here we report genome-wide identification of genes enabling accurate prediction of hybrid offspring complex traits from parents using maize grain yield as the target trait. We identified 181 ZmF1GY genes enabling prediction of maize (Zea mays L.) F₁ hybrid grain yield from parents and tested their utility and efficiency for predicting F₁ hybrid grain yields from parents using their expressions, genic SNPs, and number of favorable alleles (NFAs), respectively. The ZmF1GY genes predicted hybrid grain yields from parents at an accuracy of 0.86, presented by correlation coefficient between predicted and observed phenotypes, within an environment, 0.74 across environments, and 0.64 across populations, outperforming genomic prediction by 27-406%, 23%, and 40%, respectively. Furthermore, we identified nine of the ZmF1GY genes containing SNPs or InDels in parents that increased or decreased hybrid grain yields by 14-46%. When the NFAs of these nine ZmF1GY genes were used for hybrid grain yield prediction from parents, they predicted hybrid grain yields at an accuracy of 0.79, outperforming genomic prediction by 21% that was based on up to tens of thousands of genome-wide SNPs. These results demonstrate the feasibility of developing a gene toolkit for a species enabling gene-based breeding across environments and populations that is much more powerful and efficient than current breeding, thereby helping secure the world's food production. The methodology is applicable to all crops, livestock, and humans.

Phenomic selection in wheat breeding: prediction of the genotype-by-environment interaction in multi-environment breeding trials

Phenomic prediction of wheat grain yield and heading date in different multi-environmental trial scenarios is accurate. Modelling the genotype-by-environment interaction effect using phenomic data is a potentially low-cost complement to genomic prediction. The performance of wheat cultivars in multi-environmental trials (MET) is difficult to predict because of the genotype-by-environment interactions (G × E). Phenomic selection is supposed to be efficient for modelling the G × E effect because it accounts for non-additive effects. Here, phenomic data are near-infrared (NIR) spectra obtained from plant material. While phenomic selection has recently been shown to accurately predict wheat grain yield in single environments, its accuracy needs to be investigated for MET. We used four datasets from two winter wheat breeding programs to test and compare the predictive abilities of phenomic and genomic models for grain yield and heading date in different MET scenarios. We also compared different methods to model the G × E using different covariance matrices based on spectra. On average, phenomic and genomic prediction abilities are similar in all different MET scenarios. Better predictive abilities were obtained when G × E effects were modelled with NIR spectra than without them, and it was better to use all the spectra of all genotypes in all environments for modelling the G × E. To facilitate the implementation of phenomic prediction, we tested MET designs where the NIR spectra were measured only on the genotype-environment combinations phenotyped for the target trait. Missing spectra were predicted with a weighted multivariate ridge regression. Intermediate predictive abilities for grain yield were obtained in a sparse testing scenario and for new genotypes, which shows that phenomic selection is an efficient and practicable prediction method for dealing with G × E.

Incorporating kernelized multi-omics data improves the accuracy of genomic prediction

Background

Genomic selection (GS) has revolutionized animal and plant breeding after the first implementation via early selection before measuring phenotypes. Besides genome, transcriptome and metabolome information are increasingly considered new sources for GS. Difficulties in building the model with multi-omics data for GS and the limit of specimen availability have both delayed the progress of investigating multi-omics.

Results

We utilized the Cosine kernel to map genomic and transcriptomic data as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${n}\times {n}$$\end{document}n×n symmetric matrix (G matrix and T matrix), combined with the best linear unbiased prediction (BLUP) for GS. Here, we defined five kernel-based prediction models: genomic BLUP (GBLUP), transcriptome-BLUP (TBLUP), multi-omics BLUP (MBLUP, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\boldsymbol M=\mathrm{ratio}\times\boldsymbol G+(1-\mathrm{ratio})\times\boldsymbol T$$\end{document}M=ratio×G+(1-ratio)×T), multi-omics single-step BLUP (mssBLUP), and weighted multi-omics single-step BLUP (wmssBLUP) to integrate transcribed individuals and genotyped resource population. The predictive accuracy evaluations in four traits of the Chinese Simmental beef cattle population showed that (1) MBLUP was far preferred to GBLUP (ratio = 1.0), (2) the prediction accuracy of wmssBLUP and mssBLUP had 4.18% and 3.37% average improvement over GBLUP, (3) We also found the accuracy of wmssBLUP increased with the growing proportion of transcribed cattle in the whole resource population.

Conclusions

We concluded that the inclusion of transcriptome data in GS had the potential to improve accuracy. Moreover, wmssBLUP is accepted to be a promising alternative for the present situation in which plenty of individuals are genotyped when fewer are transcribed.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40104-022-00756-6.

De novo assembly of two chromosome-level rice genomes and bin-based QTL mapping reveal genetic diversity of grain weight trait in rice

Following the "green revolution," indica and japonica hybrid breeding has been recognized as a new breakthrough in further improving rice yields. However, heterosis-related grain weight QTLs and the basis of yield advantage among subspecies has not been well elucidated. We herein de novo assembled the chromosome level genomes of an indica/xian rice (Luohui 9) and a japonica/geng rice (RPY geng) and found that gene number differences and structural variations between these two genomes contribute to the differences in agronomic traits and also provide two different favorable allele pools to produce better derived recombinant inbred lines (RILs). In addition, we generated a high-generation (> F₁₅) population of 272 RILs from the cross between Luohui 9 and RPY geng and two testcross hybrid populations derived from the crosses of RILs and two cytoplasmic male sterile lines (YTA, indica and Z7A, japonica). Based on three derived populations, we totally identified eight 1,000-grain weight (KGW) QTLs and eight KGW heterosis loci. Of QTLs, qKGW-6.1 and qKGW-8.1 were accepted as novel KGW QTLs that have not been reported previously. Interestingly, allele genotyping results revealed that heading date related gene (Ghd8) in qKGW-8.1 and qLH-KGW-8.1, can affect grain weight in RILs and rice core accessions and may also play an important role in grain weight heterosis. Our results provided two high-quality genomes and novel gene editing targets for grain weight for future rice yield improvement project.

Predicting hybrid rice performance using AIHIB model based on artificial intelligence

Hybrid breeding is fast becoming a key instrument in plants' crop productivity. Grain yield performance of hybrids (F1) under different parental genetic features has consequently received considerable attention in the literature. The main objective of this study was to introduce a new method, known as AI_HIB under different parental genetic features using artificial intelligence (AI) techniques. In so doing, the rice cultivars TAM, KHZ, SPD, GHB, IR28, AHM, SHP and their F₁ hybrid were used. Having recorded Grain Yield (GY), Unfertile Panicle Number (UFP), Plant Height (HE), Days to Flowering (DF), Panicle Exertion (PE), Panicle Length (PL), Filled Grain Number (FG), Primary Branches Number (PBN), Flag Leaf Length (FLL), Flag Leaf Width (FLW), Flag Leaf Area (FLA), and Plant Biomass (BI) in the field, we include these features in our proposed model. When using the GA and PSO algorithm to select the features, grain yield had the highest frequency at the input of the Artificial Neural Network (ANN), Adaptive Neuro-Fuzzy Inference System (ANFIS) and Support Vector Machine (SVM) structure. The AI_HIB_ANN result revealed that the trained neural network with parental data enjoyed a good ability to predict the response of hybrid performance. Findings also reflected that the obtained MSE was low and R² value was greater than 96%. AI_HIB_SVM and AI_HIB_ANFIS showed that measuring attributes could predict number of primary branches, plant height, days to flowering and grain yield per plant with accuracies of 99%. These findings have significant implications as it presents a new promising prediction method for hybrid rice yield based on the characteristics of the parent lines by AI. These findings contribute to provide a basis for designing a smartphone application in terms of the AI_HIB_SVM and AI_HIB_ANFIS methods to easily predict hybrid performance with a high accuracy rate.

Can biochemical traits bridge the gap between genomics and plant performance? A study in rice under drought

The possibility of introducing metabolic/biochemical phenotyping to complement genomics-based predictions in breeding pipelines has been considered for years. Here we examine to what extent and under what environmental conditions metabolic/biochemical traits can effectively contribute to understanding and predicting plant performance. In this study, multivariable statistical models based on flag leaf central metabolism and oxidative stress status were used to predict grain yield (GY) performance for 271 indica rice (Oryza sativa) accessions grown in the field under well-watered and reproductive stage drought conditions. The resulting models displayed significantly higher predictability than multivariable models based on genomic data for the prediction of GY under drought (Q2 = 0.54-0.56 versus 0.35) and for stress-induced GY loss (Q2 = 0.59-0.64 versus 0.03-0.06). Models based on the combined datasets showed predictabilities similar to metabolic/biochemical-based models alone. In contrast to genetic markers, models with enzyme activities and metabolite values also quantitatively integrated the effect of physiological differences such as plant height on GY. The models highlighted antioxidant enzymes of the ascorbate-glutathione cycle and a lipid oxidation stress marker as important predictors of rice GY stability under drought at the reproductive stage, and these stress-related variables were more predictive than leaf central metabolites. These findings provide evidence that metabolic/biochemical traits can integrate dynamic cellular and physiological responses to the environment and can help bridge the gap between the genome and the phenome of crops as predictors of GY performance under drought.

Metabolomic spectra for phenotypic prediction of malting quality in spring barley

We investigated prediction of malting quality (MQ) phenotypes in different locations using metabolomic spectra, and compared the prediction ability of different models, and training population (TP) sizes. Data of five MQ traits was measured on 2667 individual plots of 564 malting spring barley lines from three years and two locations. A total of 24,018 metabolomic features (MFs) were measured on each wort sample. Two statistical models were used, a metabolomic best linear unbiased prediction (MBLUP) and a partial least squares regression (PLSR). Predictive ability within location and across locations were compared using cross-validation methods. For all traits, more than 90% of the total variance in MQ traits could be explained by MFs. The prediction accuracy increased with increasing TP size and stabilized when the TP size reached 1000. The optimal number of components considered in the PLSR models was 20. The accuracy using leave-one-line-out cross-validation ranged from 0.722 to 0.865 and using leave-one-location-out cross-validation from 0.517 to 0.817. In conclusion, the prediction accuracy of metabolomic prediction of MQ traits using MFs was high and MBLUP is better than PLSR if the training population is larger than 100. The results have significant implications for practical barley breeding for malting quality.

High-resolution bin-based linkage mapping uncovers the genetic architecture and heterosis-related loci of plant height in indica-japonica derived populations

Plant height (PH) is an important trait affecting the plant architecture, seed yield, and harvest index. However, the molecular mechanisms underlying PH heterosis remain unclear. In addition, useful PH-related genes must be urgently identified to facilitate ideal plant architecture breeding in rice (Oryza sativa L.). In the present study, to explore rice quantitative trait loci (QTLs) and heterosis-related loci of PH in rice, we developed a high-generation (>F₁₅ ) population of 272 recombinant inbred lines (RIL) from a cross of two elite varieties, Luohui 9 (indica/xian) × RPY geng (japonica/geng), and two testcross hybrid populations derived from the crosses of RILs and two cytoplasmic male sterile lines (YTA [indica] and Z7A [japonica]). Using deep resequencing data, a high-density genetic map containing 4758 bin markers was constructed, with a total map distance of 2356.41 cM. Finally, 31 PH-related QTLs for different PH component lengths or tiller numbers across five seasons were identified. Two major environment-specific PH QTLs were stably detected in Hainan (qPH-3.1) or Hubei (qPH-5.1), which have undergone significant functional alterations in rice with changes in geographical environment. Based on comparative genomics, gene function annotation, homolog identification, and existing literature (pioneering studies), candidate genes for multiple QTLs were fine-mapped, and the candidate genes qPH-3.1 and qPH-5.1 for PH were further validated using CRISPR-Cas9 gene editing. Specifically, qPH-3.1 was characterized as a pleiotropic gene, and the qPH-3.1 knockout line showed reduced PH, delayed heading, a decreased seed setting rate, and increased tiller numbers. Importantly, 10 PH heterosis-related QTLs were identified in the testcross populations, and a better-parent heterosis locus (qBPH-5.2) completely covered qPH-5.1. Furthermore, the cross results of fixed-genotype RILs verified the dominant effects of qPH-3.1 and qPH-5.1. Together, these findings further our understanding of the genetic mechanisms of PH and offer multiple highly reliable gene targets for breeding rice varieties with ideal architecture and high yield potential in the immediate future.

Phenomic Selection: A New and Efficient Alternative to Genomic Selection

Recently, it has been proposed to switch molecular markers to near-infrared (NIR) spectra for inferring relationships between individuals and further performing phenomic selection (PS), analogous to genomic selection (GS). The PS concept is similar to genomic-like omics-based (GLOB) selection, in which molecular markers are replaced by endophenotypes, such as metabolites or transcript levels, except that the phenomic information obtained for instance by near-infrared spectroscopy (NIRS ) has usually a much lower cost than other omics. Though NIRS has been routinely used in breeding for several decades, especially to deal with end-product quality traits, its use to predict other traits of interest and further make selections is new. Since the seminal paper on PS , several publications have advocated the use of spectral acquisition (including NIRS and hyperspectral imaging) in plant breeding towards PS , potentially providing a scope of what is possible. In the present chapter, we first come back to the concept of PS as originally proposed and provide a classification of selected papers related to the use of phenomics in breeding. We further provide a review of the selected literature concerning the type of technology used, the preprocessing of the spectra, and the statistical modeling to make predictions. We discuss the factors that likely affect the efficiency of PS and compare it to GS in terms of predictive ability. Finally, we propose several prospects for future work and application of PS in the context of plant breeding.

Target-oriented prioritization: targeted selection strategy by integrating organismal and molecular traits through predictive analytics in breeding

Genomic prediction in crop breeding is hindered by modeling on limited phenotypic traits. We propose an integrative multi-trait breeding strategy via machine learning algorithm, target-oriented prioritization (TOP). Using a large hybrid maize population, we demonstrate that the accuracy for identifying a candidate that is phenotypically closest to an ideotype, or target variety, achieves up to 91%. The strength of TOP is enhanced when omics level traits are included. We show that TOP enables selection of inbreds or hybrids that outperform existing commercial varieties. It improves multiple traits and accurately identifies improved candidates for new varieties, which will greatly influence breeding.

Leveraging a graft collection to develop metabolome-based trait prediction for the selection of tomato rootstocks with enhanced salt tolerance

Grafting has been demonstrated to significantly enhance the salt tolerance of crops. However, breeding efforts to develop enhanced graft combinations are hindered by knowledge-gaps as to how rootstocks mediate scion-response to salt stress. We grafted the scion of cultivated M82 onto rootstocks of 254 tomato accessions and explored the morphological and metabolic responses of grafts under saline conditions (EC = 20 dS m^-1) as compared to self-grafted M82 (SG-M82). Correlation analysis and Least Absolute Shrinkage and Selection Operator were performed to address the association between morphological diversification and metabolic perturbation. We demonstrate that grafting the same variety onto different rootstocks resulted in scion phenotypic heterogeneity and emphasized the productivity efficiency of M82 irrespective of the rootstock. Spectrophotometric analysis to test lipid oxidation showed largest variability of malondialdehyde (MDA) equivalents across the population, while the least responsive trait was the ratio of fruit fresh weight to total fresh weight (FFW/TFW). Generally, grafts showed greater values for the traits measured than SG-M82, except for branch number and wild race-originated rootstocks; the latter were associated with smaller scion growth parameters. Highly responsive and correlated metabolites were identified across the graft collection including malate, citrate, and aspartate, and their variance was partly related to rootstock origin. A group of six metabolites that consistently characterized exceptional graft response was observed, consisting of sorbose, galactose, sucrose, fructose, myo-inositol, and proline. The correlation analysis and predictive modelling, integrating phenotype- and leaf metabolite data, suggest a potential predictive relation between a set of leaf metabolites and yield-related traits.

Estimating genetic variance contributed by a quantitative trait locus: A random model approach

Detecting quantitative trait loci (QTL) and estimating QTL variances (represented by the squared QTL effects) are two main goals of QTL mapping and genome-wide association studies (GWAS). However, there are issues associated with estimated QTL variances and such issues have not attracted much attention from the QTL mapping community. Estimated QTL variances are usually biased upwards due to estimation being associated with significance tests. The phenomenon is called the Beavis effect. However, estimated variances of QTL without significance tests can also be biased upwards, which cannot be explained by the Beavis effect; rather, this bias is due to the fact that QTL variances are often estimated as the squares of the estimated QTL effects. The parameters are the QTL effects and the estimated QTL variances are obtained by squaring the estimated QTL effects. This square transformation failed to incorporate the errors of estimated QTL effects into the transformation. The consequence is biases in estimated QTL variances. To correct the biases, we can either reformulate the QTL model by treating the QTL effect as random and directly estimate the QTL variance (as a variance component) or adjust the bias by taking into account the error of the estimated QTL effect. A moment method of estimation has been proposed to correct the bias. The method has been validated via Monte Carlo simulation studies. The method has been applied to QTL mapping for the 10-week-body-weight trait from an F₂ mouse population.

One of the goals of QTL mapping and GWAS is to quantify the size of a QTL, which is measured by the QTL variance or the proportion of trait variance explained by the QTL. The effect of a QTL appears in a linear or linear mixed model as a regression coefficient and defined as a fixed effect. The estimated QTL variance in conventional QTL mapping studies takes the square of the estimated QTL effect. This is a biased estimate of QTL variance. An unbiased estimate of the QTL variance should be obtained by (1) treating the QTL effect as random and estimating the variance of the random effect or (2) adjusting the squared estimated QTL effect by the squared estimation error. We proved that the two methods are identical. We further proved that the usual R² (goodness of fit) in regression analysis is equivalent to the biased QTL heritability while the adjusted R² is equivalent to the bias corrected QTL heritability.

Accurate prediction of complex traits for individuals and offspring from parents using a simple, rapid, and efficient method for gene-based breeding in cotton and maize

Accurate, simple, rapid, and inexpensive prediction of complex traits controlled by numerous genes is paramount to enhanced plant breeding, animal breeding, and human medicine. Here we report a novel method that enables accurate, simple, and rapid prediction of complex traits of individuals or offspring from parents based on the number of favorable alleles (NFAs) of the genes controlling the objective traits. The NFAs of 226 cotton fiber length (GFL) genes and nine maize hybrid grain yield related (ZmF1GY) genes were directly used to predict cotton fiber lengths of individual plants and maize grain yields of F₁ hybrids from parents, respectively, using prediction model-based methods as controls. The NFAs of the 226 GFL genes predicted cotton fiber lengths at an accuracy of 0.85, as the model methods and outperforming genomic prediction by 82 % - 170 %. The NFAs of the nine ZmF1GY genes predicted grain yields of maize hybrids from parents at an accuracy of 0.80, outperforming genomic prediction by 67 %. Moreover, the prediction accuracies of these traits were consistent across years, environments, and eco-agricultural systems. Importantly, the accurate prediction of these traits directly using the NFAs of the genes allows breeding to be performed in greenhouse, phytotron, or off-season, without the need of the model training and validation steps essential and costly for model-based genomic or genic prediction. Therefore, this new method dramatically outperforms the current model-based genomic methods used for phenotype prediction and streamlines the process of breeding, thus promising to substantially enhance current plant and animal breeding.

Assembly of yield heterosis of an elite rice hybrid is promising by manipulating dominant quantitative trait loci

In the past, rice hybrids with strong heterosis have been obtained empirically, by developing and testing thousands of combinations. Here, we aimed to determine whether heterosis of an elite hybrid could be achieved by manipulating major quantitative trait loci. We used 202 chromosome segment substitution lines from the elite hybrid Shanyou 63 to evaluate single segment heterosis (SSH) of yield per plant and identify heterotic loci. All nine detected heterotic loci acted in a dominant fashion, and no SSH exhibited overdominance. Functional alleles of key yield-related genes Ghd7, Ghd7.1, Hd1, and GS3 were dispersed in both parents. No functional alleles of three investigated genes were expressed at higher levels in the hybrids than in the more desirable parents. A hybrid pyramiding eight heterotic loci in the female parent Zhenshan 97 background had a comparable yield to Shanyou 63 and much higher yield than Zhenshan 97. Five hybrids pyramiding eight or nine heterotic loci in the combined parental genome background showed similar yield performance to that of Shanyou 63. These results suggest that dominance underlying functional complementation is an important contributor to yield heterosis and that heterosis assembly might be successfully promised by manipulating several major dominant heterotic loci.

Phenomic selection in wheat breeding: identification and optimisation of factors influencing prediction accuracy and comparison to genomic selection

Phenomic selection is a promising alternative or complement to genomic selection in wheat breeding. Models combining spectra from different environments maximise the predictive ability of grain yield and heading date of wheat breeding lines. Phenomic selection (PS) is a recent breeding approach similar to genomic selection (GS) except that genotyping is replaced by near-infrared (NIR) spectroscopy. PS can potentially account for non-additive effects and has the major advantage of being low cost and high throughput. Factors influencing GS predictive abilities have been intensively studied, but little is known about PS. We tested and compared the abilities of PS and GS to predict grain yield and heading date from several datasets of bread wheat lines corresponding to the first or second years of trial evaluation from two breeding companies and one research institute in France. We evaluated several factors affecting PS predictive abilities including the possibility of combining spectra collected in different environments. A simple H-BLUP model predicted both traits with prediction ability from 0.26 to 0.62 and with an efficient computation time. Our results showed that the environments in which lines are grown had a crucial impact on predictive ability based on the spectra acquired and was specific to the trait considered. Models combining NIR spectra from different environments were the best PS models and were at least as accurate as GS in most of the datasets. Furthermore, a GH-BLUP model combining genotyping and NIR spectra was the best model of all (prediction ability from 0.31 to 0.73). We demonstrated also that as for GS, the size and the composition of the training set have a crucial impact on predictive ability. PS could therefore replace or complement GS for efficient wheat breeding programs.

Rice metabolic regulatory network spanning the entire life cycle

As one of the most important crops in the world, rice (Oryza sativa) is a model plant for metabolome research. Although many studies have focused on the analysis of specific tissues, the changes in metabolite abundance across the entire life cycle have not yet been determined. In this study, combining both targeted and nontargeted metabolite profiling methods, a total of 825 annotated metabolites were quantified in rice samples from different tissues covering the entire life cycle. The contents of metabolites in different tissues of rice were significantly different, with various metabolites accumulating in the plumule and radicle during seed germination. Combining these data with transcriptome data obtained from the same time period, we constructed the Rice Metabolic Regulation Network. The metabolites and co-expressed genes were further divided into 12 clusters according to their accumulation patterns, with members within each cluster displaying a uniform and clear pattern of abundance across development. Using this dataset, we established a comprehensive metabolic profile of the rice life cycle and used two independent strategies to identify novel transcription factors-namely the use of known regulatory genes as bait to screen for new networks underlying lignin metabolism and the unbiased identification of new glycerophospholipid metabolism regulators on the basis of tissue specificity. This study thus demonstrates how guilt-by-association analysis of metabolome and transcriptome data spanning the entire life cycle in cereal crops provides novel resources and tools to aid in understanding the mechanisms underlying important agronomic traits.

Re-imagining crop domestication in the era of high throughput phenomics

De novo domestication is an exciting option for increasing species diversity and ecosystem service functionality of agricultural landscapes. Genomic selection (GS), the application of genomic markers to predict phenotypic traits in a breeding population, offers the possibility of rapid genetic improvement, making GS especially attractive for modifying traits of long-lived species. However, for some wild species just entering the domestication pipeline, especially those with large and complex genomes, a lack of funding and/or prior genome characterization, GS is often out of reach. High throughput phenomics has the potential to augment traditional pedigree selection, reduce costs and amplify impacts of genomic selection, and even create new predictive selection approaches independent of sequencing or pedigrees.

Prediction of heterosis in the recent rapeseed (Brassica napus) polyploid by pairing parental nucleotide sequences

The utilization of heterosis is a successful strategy in increasing yield for many crops. However, it consumes tremendous manpower to test the combining ability of the parents in fields. Here, we applied the genomic-selection (GS) strategy and developed models that significantly increase the predictability of heterosis by introducing the concept of a regional parental genetic-similarity index (PGSI) and reducing dimension in the calculation matrix in a machine-learning approach. Overall, PGSI negatively affected grain yield and several other traits but positively influenced the thousand-seed weight of the hybrids. It was found that the C subgenome of rapeseed had a greater impact on heterosis than the A subgenome. We drew maps with overviews of quantitative-trait loci that were responsible for the heterosis (h-QTLs) of various agronomic traits. Identifications and annotations of genes underlying high impacting h-QTLs were provided. Using models that we elaborated, combining abilities between an Ogu-CMS-pool member and a potential restorer can be simulated in silico, sidestepping laborious work, such as testing crosses in fields. The achievements here provide a case of heterosis prediction in polyploid genomes with relatively large genome sizes.

Oilseed rape (Brassica napus) is of significant economic interest worldwide, providing high-quality oil with excellent health-promoting properties. It represents an excellent model of a successful recent polyploid that rapidly became an important crop worldwide. The utilization of hybridization, leading to hybrid vigor, or heterosis, is a successful strategy in increasing yield and vigor for many field crops including rapeseed (Brassica napus). However, the procedure of using classical breeding methods remains slow and laborious, illustrating the need for predictive and innovative methods. Here, we have achieved a significant breakthrough by using genome selection and significantly advanced models to predict the heterosis by pairing genome-wide nucleotides of parents. We provided maps with overviews of quantitative trait loci that were responsible for the heterosis of various agronomic traits. The research used deep resequencing (>30x) data of the entire polyploidy rapeseed genome, providing a successful case for the prediction of heterosis in polyploid genomes with relatively large genome sizes. Moreover, we provided the genetic information (SNPs) of 1007 core accessions of this species in the public domain for testing combinations with high heterosis using our predicting model for rapeseed breeders all over the world.

The metabolomic landscape of rice heterosis highlights pathway biomarkers for predicting complex phenotypes

Understanding the molecular mechanisms underlying complex phenotypes requires systematic analyses of complicated metabolic networks and contributes to improvements in the breeding efficiency of staple cereal crops and diagnostic accuracy for human diseases. Here, we selected rice (Oryza sativa) heterosis as a complex phenotype and investigated the mechanisms of both vegetative and reproductive traits using an untargeted metabolomics strategy. Heterosis-associated analytes were identified, and the overlapping analytes were shown to underlie the association patterns for six agronomic traits. The heterosis-associated analytes of four yield components and plant height collectively contributed to yield heterosis, and the degree of contribution differed among the five traits. We performed dysregulated network analyses of the high- and low-better parent heterosis hybrids and found multiple types of metabolic pathways involved in heterosis. The metabolite levels of the significantly enriched pathways (especially those from amino acid and carbohydrate metabolism) were predictive of yield heterosis (area under the curve = 0.907 with 10 features), and the predictability of these pathway biomarkers was validated with hybrids across environments and populations. Our findings elucidate the metabolomic landscape of rice heterosis and highlight the potential application of pathway biomarkers in achieving accurate predictions of complex phenotypes.

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.

Genomic Prediction and the Practical Breeding of 12 Quantitative-Inherited Traits in Cucumber (Cucumis sativus L.)

Genomic prediction is an effective way for predicting complex traits, and it is becoming more essential in horticultural crop breeding. In this study, we applied genomic prediction in the breeding of cucumber plants. Eighty-one cucumber inbred lines were genotyped and 16,662 markers were identified to represent the genetic background of cucumber. Two populations, namely, diallel cross population and North Carolina II population, having 268 combinations in total were constructed from 81 inbred lines. Twelve cucumber commercial traits of these two populations in autumn 2018, spring 2019, and spring 2020 were collected for model training. General combining ability (GCA) models under five-fold cross-validation and cross-population validation were applied to model validation. Finally, the GCA performance of 81 inbred lines was estimated. Our results showed that the predictive ability for 12 traits ranged from 0.38 to 0.95 under the cross-validation strategy and ranged from -0.38 to 0.88 under the cross-population strategy. Besides, GCA models containing non-additive effects had significantly better performance than the pure additive GCA model for most of the investigated traits. Furthermore, there were a relatively higher proportion of additive-by-additive genetic variance components estimated by the full GCA model, especially for lower heritability traits, but the proportion of dominant genetic variance components was relatively small and stable. Our findings concluded that a genomic prediction protocol based on the GCA model theoretical framework can be applied to cucumber breeding, and it can also provide a reference for the single-cross breeding system of other crops.

PIP-SNP: a pipeline for processing SNP data featured as linkage disequilibrium bin mapping, genotype imputing and marker synthesizing

Genome-wide association study data analyses often face two significant challenges: (i) high dimensionality of single-nucleotide polymorphism (SNP) genotypes and (ii) imputation of missing values. SNPs are not independent due to physical linkage and natural selection. The correlation of nearby SNPs is known as linkage disequilibrium (LD), which can be used for LD conceptual SNP bin mapping, missing genotype inferencing and SNP dimension reduction. We used a stochastic process to describe the SNP signals and proposed two types of autocorrelations to measure nearby SNPs' information redundancy. Based on the calculated autocorrelation coefficients, we constructed LD bins. We adopted a k-nearest neighbors algorithm (kNN) to impute the missing genotypes. We proposed several novel methods to find the optimal synthetic marker to represent the SNP bin. We also proposed methods to evaluate the information loss or information conservation between using the original genome-wide markers and using dimension-reduced synthetic markers. Our performance assessments on the real-life SNP data from a rice recombinant inbred line (RIL) population and a rice HapMap project show that the new methods produce satisfactory results. We implemented these functional modules in C/C++ and streamlined them into a web-based pipeline named PIP-SNP (https://bioinfo.noble.org/PIP_SNP/) for processing SNP data.

Molecular basis of heterosis and related breeding strategies reveal its importance in vegetable breeding

Heterosis has historically been exploited in plants; however, its underlying genetic mechanisms and molecular basis remain elusive. In recent years, due to advances in molecular biotechnology at the genome, transcriptome, proteome, and epigenome levels, the study of heterosis in vegetables has made significant progress. Here, we present an extensive literature review on the genetic and epigenetic regulation of heterosis in vegetables. We summarize six hypotheses to explain the mechanism by which genes regulate heterosis, improve upon a possible model of heterosis that is triggered by epigenetics, and analyze previous studies on quantitative trait locus effects and gene actions related to heterosis based on analyses of differential gene expression in vegetables. We also discuss the contributions of yield-related traits, including flower, fruit, and plant architecture traits, during heterosis development in vegetables (e.g., cabbage, cucumber, and tomato). More importantly, we propose a comprehensive breeding strategy based on heterosis studies in vegetables and crop plants. The description of the strategy details how to obtain F1 hybrids that exhibit heterosis based on heterosis prediction, how to obtain elite lines based on molecular biotechnology, and how to maintain heterosis by diploid seed breeding and the selection of hybrid simulation lines that are suitable for heterosis research and utilization in vegetables. Finally, we briefly provide suggestions and perspectives on the role of heterosis in the future of vegetable breeding.

Improving Genomic Prediction Using High-Dimensional Secondary Phenotypes

In the past decades, genomic prediction has had a large impact on plant breeding. Given the current advances of high-throughput phenotyping and sequencing technologies, it is increasingly common to observe a large number of traits, in addition to the target trait of interest. This raises the important question whether these additional or “secondary” traits can be used to improve genomic prediction for the target trait. With only a small number of secondary traits, this is known to be the case, given sufficiently high heritabilities and genetic correlations. Here we focus on the more challenging situation with a large number of secondary traits, which is increasingly common since the arrival of high-throughput phenotyping. In this case, secondary traits are usually incorporated through additional relatedness matrices. This approach is however infeasible when secondary traits are not measured on the test set, and cannot distinguish between genetic and non-genetic correlations. An alternative direction is to extend the classical selection indices using penalized regression. So far, penalized selection indices have not been applied in a genomic prediction setting, and require plot-level data in order to reliably estimate genetic correlations. Here we aim to overcome these limitations, using two novel approaches. Our first approach relies on a dimension reduction of the secondary traits, using either penalized regression or random forests (LS-BLUP/RF-BLUP). We then compute the bivariate GBLUP with the dimension reduction as secondary trait. For simulated data (with available plot-level data), we also use bivariate GBLUP with the penalized selection index as secondary trait (SI-BLUP). In our second approach (GM-BLUP), we follow existing multi-kernel methods but replace secondary traits by their genomic predictions, with the advantage that genomic prediction is also possible when secondary traits are only measured on the training set. For most of our simulated data, SI-BLUP was most accurate, often closely followed by RF-BLUP or LS-BLUP. In real datasets, involving metabolites in Arabidopsis and transcriptomics in maize, no method could substantially improve over univariate prediction when secondary traits were only available on the training set. LS-BLUP and RF-BLUP were most accurate when secondary traits were available also for the test set.

Prediction of complex phenotypes using the Drosophila melanogaster metabolome

Understanding the genotype-phenotype map and how variation at different levels of biological organization is associated are central topics in modern biology. Fast developments in sequencing technologies and other molecular omic tools enable researchers to obtain detailed information on variation at DNA level and on intermediate endophenotypes, such as RNA, proteins and metabolites. This can facilitate our understanding of the link between genotypes and molecular and functional organismal phenotypes. Here, we use the Drosophila melanogaster Genetic Reference Panel and nuclear magnetic resonance (NMR) metabolomics to investigate the ability of the metabolome to predict organismal phenotypes. We performed NMR metabolomics on four replicate pools of male flies from each of 170 different isogenic lines. Our results show that metabolite profiles are variable among the investigated lines and that this variation is highly heritable. Second, we identify genes associated with metabolome variation. Third, using the metabolome gave better prediction accuracies than genomic information for four of five quantitative traits analyzed. Our comprehensive characterization of population-scale diversity of metabolomes and its genetic basis illustrates that metabolites have large potential as predictors of organismal phenotypes. This finding is of great importance, e.g., in human medicine, evolutionary biology and animal and plant breeding.

Multi-omics-based prediction of hybrid performance in canola

Complementing or replacing genetic markers with transcriptomic data and use of reproducing kernel Hilbert space regression based on Gaussian kernels increases hybrid prediction accuracies for complex agronomic traits in canola. In plant breeding, hybrids gained particular importance due to heterosis, the superior performance of offspring compared to their inbred parents. Since the development of new top performing hybrids requires labour-intensive and costly breeding programmes, including testing of large numbers of experimental hybrids, the prediction of hybrid performance is of utmost interest to plant breeders. In this study, we tested the effectiveness of hybrid prediction models in spring-type oilseed rape (Brassica napus L./canola) employing different omics profiles, individually and in combination. To this end, a population of 950 F₁ hybrids was evaluated for seed yield and six other agronomically relevant traits in commercial field trials at several locations throughout Europe. A subset of these hybrids was also evaluated in a climatized glasshouse regarding early biomass production. For each of the 477 parental rapeseed lines, 13,201 single nucleotide polymorphisms (SNPs), 154 primary metabolites, and 19,479 transcripts were determined and used as predictive variables. Both, SNP markers and transcripts, effectively predict hybrid performance using (genomic) best linear unbiased prediction models (gBLUP). Compared to models using pure genetic markers, models incorporating transcriptome data resulted in significantly higher prediction accuracies for five out of seven agronomic traits, indicating that transcripts carry important information beyond genomic data. Notably, reproducing kernel Hilbert space regression based on Gaussian kernels significantly exceeded the predictive abilities of gBLUP models for six of the seven agronomic traits, demonstrating its potential for implementation in future canola breeding programmes.

Translating insights from the seed metabolome into improved prediction for lipid-composition traits in oat (Avena sativa L.)

Oat (Avena sativa L.) seed is a rich resource of beneficial lipids, soluble fiber, protein, and antioxidants, and is considered a healthful food for humans. Little is known regarding the genetic controllers of variation for these compounds in oat seed. We characterized natural variation in the mature seed metabolome using untargeted metabolomics on 367 diverse lines and leveraged this information to improve prediction for seed quality traits. We used a latent factor approach to define unobserved variables that may drive covariance among metabolites. One hundred latent factors were identified, of which 21% were enriched for compounds associated with lipid metabolism. Through a combination of whole-genome regression and association mapping, we show that latent factors that generate covariance for many metabolites tend to have a complex genetic architecture. Nonetheless, we recovered significant associations for 23% of the latent factors. These associations were used to inform a multi-kernel genomic prediction model, which was used to predict seed lipid and protein traits in two independent studies. Predictions for 8 of the 12 traits were significantly improved compared to genomic best linear unbiased prediction when this prediction model was informed using associations from lipid-enriched factors. This study provides new insights into variation in the oat seed metabolome and provides genomic resources for breeders to improve selection for health-promoting seed quality traits. More broadly, we outline an approach to distill high-dimensional "omics" data to a set of biologically meaningful variables and translate inferences on these data into improved breeding decisions.

Genome-wide association study on metabolite accumulation in a wild barley NAM population reveals natural variation in sugar metabolism

Metabolites play a key role in plants as they are routing plant developmental processes and are involved in biotic and abiotic stress responses. Their analysis can offer important information on the underlying processes. Regarding plant breeding, metabolite concentrations can be used as biomarkers instead of or in addition to genetic markers to predict important phenotypic traits (metabolic prediction). In this study, we applied a genome-wide association study (GWAS) in a wild barley nested association mapping (NAM) population to identify metabolic quantitative trait loci (mQTL). A set of approximately 130 metabolites, measured at early and late sampling dates, was analysed. For four metabolites from the early and six metabolites from the late sampling date significant mQTL (grouped as 19 mQTL for the early and 25 mQTL for the late sampling date) were found. Interestingly, all of those metabolites could be classified as sugars. Sugars are known to be involved in signalling, plant growth and plant development. Sugar-related genes, encoding mainly sugar transporters, have been identified as candidate genes for most of the mQTL. Moreover, several of them co-localized with known flowering time genes like Ppd-H1, HvELF3, Vrn-H1, Vrn-H2 and Vrn-H3, hinting on the known role of sugars in flowering. Furthermore, numerous disease resistance-related genes were detected, pointing to the signalling function of sugars in plant resistance. An mQTL on chromosome 1H in the region of 13 Mbp to 20 Mbp stood out, that alone explained up to 65% of the phenotypic variation of a single metabolite. Analysis of family-specific effects within the diverse NAM population showed the available natural genetic variation regarding sugar metabolites due to different wild alleles. The study represents a step towards a better understanding of the genetic components of metabolite accumulation, especially sugars, thereby linking them to biological functions in barley.

Machine learning approaches for crop improvement: Leveraging phenotypic and genotypic big data

Highly efficient and accurate selection of elite genotypes can lead to dramatic shortening of the breeding cycle in major crops relevant for sustaining present demands for food, feed, and fuel. In contrast to classical approaches that emphasize the need for resource-intensive phenotyping at all stages of artificial selection, genomic selection dramatically reduces the need for phenotyping. Genomic selection relies on advances in machine learning and the availability of genotyping data to predict agronomically relevant phenotypic traits. Here we provide a systematic review of machine learning approaches applied for genomic selection of single and multiple traits in major crops in the past decade. We emphasize the need to gather data on intermediate phenotypes, e.g. metabolite, protein, and gene expression levels, along with developments of modeling techniques that can lead to further improvements of genomic selection. In addition, we provide a critical view of factors that affect genomic selection, with attention to transferability of models between different environments. Finally, we highlight the future aspects of integrating high-throughput molecular phenotypic data from omics technologies with biological networks for crop improvement.

Incorporation of parental phenotypic data into multi-omic models improves prediction of yield-related traits in hybrid rice

Hybrid breeding has been shown to effectively increase rice productivity. However, identifying desirable hybrids out of numerous potential combinations is a daunting challenge. Genomic selection holds great promise for accelerating hybrid breeding by enabling early selection before phenotypes are measured. With the recent advances in multi-omic technologies, hybrid prediction based on transcriptomic and metabolomic data has received increasing attention. However, the current omic-based hybrid prediction has ignored parental phenotypic information, which is of fundamental importance in plant breeding. In this study, we integrated parental phenotypic information into various multi-omic prediction models applied in hybrid breeding of rice and compared the predictabilities of 15 combinations from four sets of predictors from the parents, that is genome, transcriptome, metabolome and phenome. The predictability for each combination was evaluated using the best linear unbiased prediction and a modified fast HAT method. We found significant interactions between predictors and traits in predictability, but joint prediction with various combinations of the predictors significantly improved predictability relative to prediction of any single source omic data for each trait investigated. Incorporation of parental phenotypic data into various omic predictors increased the predictability, averagely by 13.6%, 54.5%, 19.9% and 8.3%, for grain yield, number of tillers per plant, number of grains per panicle and 1000 grain weight, respectively. Among nine models of incorporating parental traits, the AD-All model was the most effective one. This novel strategy of incorporating parental phenotypic data into multi-omic prediction is expected to improve hybrid breeding progress, especially with the development of high-throughput phenotyping technologies.

Deshrinking ridge regression for genome-wide association studies

MOTIVATION

Genome-wide association studies (GWAS) are still the primary steps toward gene discovery. The urgency is more obvious in the big data era when GWAS are conducted simultaneously for thousand traits, e.g. transcriptomic and metabolomic traits. Efficient mixed model association (EMMA) and genome-wide efficient mixed model association (GEMMA) are the widely used methods for GWAS. An algorithm with high computational efficiency is badly needed. It is interesting to note that the test statistics of the ordinary ridge regression (ORR) have the same patterns across the genome as those obtained from the EMMA method. However, ORR has never been used for GWAS due to its severe shrinkage on the estimated effects and the test statistics.

RESULTS

We introduce a degree of freedom for each marker effect obtained from ORR and use it to deshrink both the estimated effect and the standard error so that the Wald test of ORR is brought back to the same level as that of EMMA. The new method is called deshrinking ridge regression (DRR). By evaluating the methods under three different model sizes (small, medium and large), we demonstrate that DRR is more generalized for all model sizes than EMMA, which only works for medium and large models. Furthermore, DRR detect all markers in a simultaneous manner instead of scanning one marker at a time. As a result, the computational time complexity of DRR is much simpler than EMMA and about m (number of genetic variants) times simpler than that of GEMMA when the sample size is way smaller than the number of markers.

CONTACT

shizhong.xu@ucr.edu.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Genomic predictions for enteric methane production are improved by metabolome and microbiome data in sheep (Ovis aries)

Methane production from rumen methanogenesis contributes approximately 71% of greenhouse gas emissions from the agricultural sector. This study has performed genomic predictions for methane production from 99 sheep across 3 yr using a residual methane phenotype that is log methane yield corrected for live weight, rumen volume, and feed intake. Using genomic relationships, the prediction accuracies (as determined by the correlation between predicted and observed residual methane production) ranged from 0.058 to 0.220 depending on the time point being predicted. The best linear unbiased prediction algorithm was then applied to relationships between animals that were built on the rumen metabolome and microbiome. Prediction accuracies for the metabolome-based relationships for the two available time points were 0.254 and 0.132; the prediction accuracy for the first microbiome time point was 0.142. The second microbiome time point could not successfully predict residual methane production. When the metabolomic relationships were added to the genomic relationships, the accuracy of predictions increased to 0.274 (from 0.201 when only the genomic relationship was used) and 0.158 (from 0.081 when only the genomic relationship was used) for the two time points, respectively. When the microbiome relationships from the first time point were added to the genomic relationships, the maximum prediction accuracy increased to 0.247 (from 0.216 when only the genomic relationship was used), which was achieved by giving the genomic relationships 10 times more weighting than the microbiome relationships. These accuracies were higher than the genomic, metabolomic, and microbiome relationship matrixes achieved alone when identical sets of animals were used.

Boosting predictabilities of agronomic traits in rice using bivariate genomic selection

The multivariate genomic selection (GS) models have not been adequately studied and their potential remains unclear. In this study, we developed a highly efficient bivariate (2D) GS method and demonstrated its significant advantages over the univariate (1D) rival methods using a rice dataset, where four traditional traits (i.e. yield, 1000-grain weight, grain number and tiller number) as well as 1000 metabolomic traits were analyzed. The novelty of the method is the incorporation of the HAT methodology in the 2D BLUP GS model such that the computational efficiency has been dramatically increased by avoiding the conventional cross-validation. The results indicated that (1) the 2D BLUP-HAT GS analysis generally produces higher predictabilities for two traits than those achieved by the analysis of individual traits using 1D GS model, and (2) selected metabolites may be utilized as ancillary traits in the new 2D BLUP-HAT GS method to further boost the predictability of traditional traits, especially for agronomically important traits with low 1D predictabilities.

Metabolomics analysis and metabolite-agronomic trait associations using kernels of wheat (Triticum aestivum) recombinant inbred lines

Plants produce numerous metabolites that are important for their development and growth. However, the genetic architecture of the wheat metabolome has not been well studied. Here, utilizing a high-density genetic map, we conducted a comprehensive metabolome study via widely targeted LC-MS/MS to analyze the wheat kernel metabolism. We further combined agronomic traits and dissected the genetic relationship between metabolites and agronomic traits. In total, 1260 metabolic features were detected. Using linkage analysis, 1005 metabolic quantitative trait loci (mQTLs) were found distributed unevenly across the genome. Twenty-four candidate genes were found to modulate the levels of different metabolites, of which two were functionally annotated by in vitro analysis to be involved in the synthesis and modification of flavonoids. Combining the correlation analysis of metabolite-agronomic traits with the co-localization of methylation quantitative trait locus (mQTL) and phenotypic QTL (pQTL), genetic relationships between the metabolites and agronomic traits were uncovered. For example, a candidate was identified using correlation and co-localization analysis that may manage auxin accumulation, thereby affecting number of grains per spike (NGPS). Furthermore, metabolomics data were used to predict the performance of wheat agronomic traits, with metabolites being found that provide strong predictive power for NGPS and plant height. This study used metabolomics and association analysis to better understand the genetic basis of the wheat metabolism which will ultimately assist in wheat breeding.

Use of multiple traits genomic prediction, genotype by environment interactions and spatial effect to improve prediction accuracy in yield data

Genomic selection has been extensively implemented in plant breeding schemes. Genomic selection incorporates dense genome-wide markers to predict the breeding values for important traits based on information from genotype and phenotype records on traits of interest in a reference population. To date, most relevant investigations have been performed using single trait genomic prediction models (STGP). However, records for several traits at once are usually documented for breeding lines in commercial breeding programs. By incorporating benefits from genetic characterizations of correlated phenotypes, multiple trait genomic prediction (MTGP) may be a useful tool for improving prediction accuracy in genetic evaluations. The objective of this study was to test whether the use of MTGP and including proper modeling of spatial effects can improve the prediction accuracy of breeding values in commercial barley and wheat breeding lines. We genotyped 1,317 spring barley and 1,325 winter wheat lines from a commercial breeding program with the Illumina 9K barley and 15K wheat SNP-chip (respectively) and phenotyped them across multiple years and locations. Results showed that the MTGP approach increased correlations between future performance and estimated breeding value of yields by 7% in barley and by 57% in wheat relative to using the STGP approach for each trait individually. Analyses combining genomic data, pedigree information, and proper modeling of spatial effects further increased the prediction accuracy by 4% in barley and 3% in wheat relative to the model using genomic relationships only. The prediction accuracy for yield in wheat and barley yield trait breeding, were improved by combining MTGP and spatial effects in the model.

Metabolome Profiling Supports the Key Role of the Spike in Wheat Yield Performance

Although the relevance of spike bracts in stress acclimation and contribution to wheat yield was recently revealed, the metabolome of this organ and its response to water stress is still unknown. The metabolite profiles of flag leaves, glumes and lemmas were characterized under contrasting field water regimes in five durum wheat cultivars. Water conditions during growth were characterized through spectral vegetation indices, canopy temperature and isotope composition. Spike bracts exhibited better coordination of carbon and nitrogen metabolisms than the flag leaves in terms of photorespiration, nitrogen assimilation and respiration paths. This coordination facilitated an accumulation of organic and amino acids in spike bracts, especially under water stress. The metabolomic response to water stress also involved an accumulation of antioxidant and drought tolerance related sugars, particularly in the spikes. Furthermore, certain cell wall, respiratory and protective metabolites were associated with genotypic outperformance and yield stability. In addition, grain yield was strongly predicted by leaf and spike bracts metabolomes independently. This study supports the role of the spike as a key organ during wheat grain filling, particularly under stress conditions and provides relevant information to explore new ways to improve wheat productivity including potential biomarkers for yield prediction.

Biomarkers for grain yield stability in rice under drought stress

Crop yield stability requires an attenuation of the reduction of yield losses caused by environmental stresses such as drought. Using a combination of metabolomics and high-throughput colorimetric assays, we analysed central metabolism and oxidative stress status in the flag leaf of 292 indica rice (Oryza sativa) accessions. Plants were grown in the field and were, at the reproductive stage, exposed to either well-watered or drought conditions to identify the metabolic processes associated with drought-induced grain yield loss. Photorespiration, protein degradation, and nitrogen recycling were the main processes involved in the drought-induced leaf metabolic reprogramming. Molecular markers of drought tolerance and sensitivity in terms of grain yield were identified using a multivariate model based on the values of the metabolites and enzyme activities across the population. The model highlights the central role of the ascorbate-glutathione cycle, particularly dehydroascorbate reductase, in minimizing drought-induced grain yield loss. In contrast, malondialdehyde was an accurate biomarker for grain yield loss, suggesting that drought-induced lipid peroxidation is the major constraint under these conditions. These findings highlight new breeding targets for improved rice grain yield stability under drought.

Accurate Prediction of a Quantitative Trait Using the Genes Controlling the Trait for Gene-Based Breeding in Cotton

Accurate phenotype prediction of quantitative traits is paramount to enhanced plant research and breeding. Here, we report the accurate prediction of cotton fiber length, a typical quantitative trait, using 474 cotton (Gossypium ssp.) fiber length (GFL) genes and nine prediction models. When the SNPs/InDels contained in 226 of the GFL genes or the expressions of all 474 GFL genes was used for fiber length prediction, a prediction accuracy of r = 0.83 was obtained, approaching the maximally possible prediction accuracy of a quantitative trait. This has improved by 116%, the prediction accuracies of the fiber length thus far achieved for genomic selection using genome-wide random DNA markers. Moreover, analysis of the GFL genes identified 125 of the GFL genes that are key to accurate prediction of fiber length, with which a prediction accuracy similar to that of all 474 GFL genes was obtained. The fiber lengths of the plants predicted with expressions of the 125 key GFL genes were significantly correlated with those predicted with the SNPs/InDels of the above 226 SNP/InDel-containing GFL genes (r = 0.892, P = 0.000). The prediction accuracies of fiber length using both genic datasets were highly consistent across environments or generations. Finally, we found that a training population consisting of 100-120 plants was sufficient to train a model for accurate prediction of a quantitative trait using the genes controlling the trait. Therefore, the genes controlling a quantitative trait are capable of accurately predicting its phenotype, thereby dramatically improving the ability, accuracy, and efficiency of phenotype prediction and promoting gene-based breeding in cotton and other species.

Metabolome-based prediction of yield heterosis contributes to the breeding of elite rice

Metabolite profiles obtained from parental seedlings can predict yield heterosis of rice hybrids with correctly dissected population structure across different growth conditions.

Improvement of the breeding efficiencies of heterotic crops adaptive to different conditions can mitigate the food shortage crisis due to overpopulation and climate change. To date, diverse molecular markers have been used to guide field phenotypic selection, whereas accurate predictions of complex heterotic traits are rarely reported. Here, we present a practical metabolome-based strategy for predicting yield heterosis in rice. The dissection of population structure based on untargeted metabolite profiles as the initial critical step in multivariate modeling performed better than the screening of predictive variables. Then the assessment of each predictive variable’s contribution to predictive models according to all latent factors was more precise than the conventional first one. Metabolites belonging to specific pathways were closely associated with yield heterosis, and the up-regulation of galactose metabolism promoted robust yield heterosis in hybrids under different growth conditions. Our study demonstrates that metabolome-based predictive models with correctly dissected population structure and screened predictive variables can facilitate accurate predictions of yield heterosis and have great potential for establishing molecular marker–based precision breeding programs.

A directed learning strategy integrating multiple omic data improves genomic prediction

Genomic prediction (GP) aims to construct a statistical model for predicting phenotypes using genome-wide markers and is a promising strategy for accelerating molecular plant breeding. However, current progress of phenotype prediction using genomic data alone has reached a bottleneck, and previous studies on transcriptomic and metabolomic predictions ignored genomic information. Here, we designed a novel strategy of GP called multilayered least absolute shrinkage and selection operator (MLLASSO) by integrating multiple omic data into a single model that iteratively learns three layers of genetic features (GFs) supervised by observed transcriptome and metabolome. Significantly, MLLASSO learns higher order information of gene interactions, which enables us to achieve a significant improvement of predictability of yield in rice from 0.1588 (GP alone) to 0.2451 (MLLASSO). In the prediction of the first two layers, some genes were found to be genetically predictable genes (GPGs) as their expressions were accurately predicted with genetic markers. Interestingly, we made three dramatic discoveries for the GPGs: (i) GPGs are good predictors for highly complex traits like yield; (ii) GPGs are mostly eQTL genes (cis or trans); and (iii) trait-related transcriptional factor families are enriched in GPGs. These findings support the notion that learned GFs not only are good predictors for traits but also have specific biological implications regarding regulation of gene expressions. To differentiate the new method from conventional GP models, we called MLLASSO a directed learning strategy supervised by intermediate omic data. This new prediction model appears to be more reliable and more robust than conventional GP models.

Effect of Imidacloprid Uptake from Contaminated Soils on Vegetable Growth

In the present study, the effect of imidacloprid uptake from contaminated soils on the growth of leaf vegetable Shanghaiqing was investigated. The result showed that during 35-day exposure, the concentration of imidacloprid (IMI) was in the order of vegetable shoots > vegetable roots > soil, indicating that IMI was more readily concentrated in vegetable shoots than in roots. Moreover, the biomass of IMI-treated vegetable shoots was comparable to that of the controls with early exposure, but was higher than that of the controls after 7-day exposure, showing that the test concentration of IMI could stimulate vegetable growth. The plant metabolic analysis of vegetable shoots using LC-QTOF/MS revealed that IMI may cause oxidative stress to the plant shoots with early exposure; however, the stressful situation of IMI seems to be relieved with the increase of some substances (such as spermidine and phenylalanine) with late exposure. Moreover, the upregulation of N-rich amino acids (glutamine, aspartate, and arginine) suggested that the process of fixing inorganic nitrogen in the plant should be enhanced, possibly contributing to enhanced growth rates. Additionally, four IMI's metabolites were identified by using MS-FINDER software, and the distribution of three metabolites in vegetable tissues was compared.