Improving Genomic Prediction Using High-Dimensional Secondary Phenotypes

Bivariate GWA mapping reveals associations between aliphatic glucosinolates and plant responses to thrips and heat stress

Although plants harbor a huge phytochemical diversity, only a fraction of plant metabolites is functionally characterized. In this work, we aimed to identify the genetic basis of metabolite functions during harsh environmental conditions in Arabidopsis thaliana. With machine learning algorithms we predicted stress-specific metabolomes for 23 (a)biotic stress phenotypes of 300 natural Arabidopsis accessions. The prediction models identified several aliphatic glucosinolates (GLSs) and their breakdown products to be implicated in responses to heat stress in siliques and herbivory by Western flower thrips, Frankliniella occidentalis. Bivariate GWA mapping of the metabolome predictions and their respective (a)biotic stress phenotype revealed genetic associations with MAM, AOP, and GS-OH, all three involved in aliphatic GSL biosynthesis. We, therefore, investigated thrips herbivory on AOP, MAM, and GS-OH loss-of-function and/or overexpression lines. Arabidopsis accessions with a combination of MAM2 and AOP3, leading to 3-hydroxypropyl dominance, suffered less from thrips feeding damage. The requirement of MAM2 for this effect could, however, not be confirmed with an introgression line of ecotypes Cvi and Ler, most likely due to other, unknown susceptibility factors in the Ler background. However, AOP2 and GS-OH, adding alkenyl or hydroxy-butenyl groups, respectively, did not have major effects on thrips feeding. Overall, this study illustrates the complex implications of aliphatic GSL diversity in plant responses to heat stress and a cell-content-feeding herbivore.

Dissecting the effect of heat stress on durum wheat under field conditions

Introduction

Heat stress negatively affects wheat production in several ways, mainly by reducing growth rate, photosynthetic capacity and reducing spike fertility. Modeling stress response means analyzing simultaneous relationships among traits affecting the whole plant response and determinants of grain yield. The aim of this study was to dissect the diverse impacts of heat stress on key yield traits and to identify the most promising sources of alleles for heat tolerance.

Methods

We evaluated a diverse durum wheat panel of 183 cultivars and breeding lines from worldwide, for their response to long-term heat stress under field conditions (HS) with respect to non stress conditions (NS), considering phenological traits, grain yield (GY) and its components as a function of the timing of heat stress and climatic covariates. We investigated the relationships among plant and environmental variables by means of a structural equation model (SEM) and Genetic SEM (GSEM).

Results

Over two years of experiments at CENEB, CIMMYT, the effects of HS were particularly pronounced for the normalized difference vegetation index, NDVI (-51.3%), kernel weight per spike, KWS (-40.5%), grain filling period, GFP (-38.7%), and GY (-56.6%). Average temperatures around anthesis were negatively correlated with GY, thousand kernel weight TKW and test weight TWT, but also with spike density, a trait determined before heading/anthesis. Under HS, the correlation between the three major determinants of GY, i.e., fertile spike density, spike fertility and kernel size, were of noticeable magnitude. NDVI measured at medium milk-soft dough stage under HS was correlated with both spike fertility and grain weight while under NS it was less predictive of grain weight but still highly correlated with spike fertility. GSEM modeling suggested that the causal model of performance under HS directly involves genetic effects on GY, NDVI, KWS and HD.

Discussion

We identified consistently suitable sources of genetic resistance to heat stress to be used in different durum wheat pre-breeding programs. Among those, Desert Durums and CIMMYT'80 germplasm showed the highest degree of adaptation and capacity to yield under high temperatures and can be considered as a valuable source of alleles for adaptation to breed new HS resilient cultivars.

Recent advances in artificial intelligence, mechanistic models, and speed breeding offer exciting opportunities for precise and accelerated genomics-assisted breeding

Given the challenges of population growth and climate change, there is an urgent need to expedite the development of high-yielding stress-tolerant crop cultivars. While traditional breeding methods have been instrumental in ensuring global food security, their efficiency, precision, and labour intensiveness have become increasingly inadequate to address present and future challenges. Fortunately, recent advances in high-throughput phenomics and genomics-assisted breeding (GAB) provide a promising platform for enhancing crop cultivars with greater efficiency. However, several obstacles must be overcome to optimize the use of these techniques in crop improvement, such as the complexity of phenotypic analysis of big image data. In addition, the prevalent use of linear models in genome-wide association studies (GWAS) and genomic selection (GS) fails to capture the nonlinear interactions of complex traits, limiting their applicability for GAB and impeding crop improvement. Recent advances in artificial intelligence (AI) techniques have opened doors to nonlinear modelling approaches in crop breeding, enabling the capture of nonlinear and epistatic interactions in GWAS and GS and thus making this variation available for GAB. While statistical and software challenges persist in AI-based models, they are expected to be resolved soon. Furthermore, recent advances in speed breeding have significantly reduced the time (3-5-fold) required for conventional breeding. Thus, integrating speed breeding with AI and GAB could improve crop cultivar development within a considerably shorter timeframe while ensuring greater accuracy and efficiency. In conclusion, this integrated approach could revolutionize crop breeding paradigms and safeguard food production in the face of population growth and climate change.

Integrating and optimizing genomic, weather, and secondary trait data for multiclass classification

Modern plant breeding programs collect several data types such as weather, images, and secondary or associated traits besides the main trait (e.g., grain yield). Genomic data is high-dimensional and often over-crowds smaller data types when naively combined to explain the response variable. There is a need to develop methods able to effectively combine different data types of differing sizes to improve predictions. Additionally, in the face of changing climate conditions, there is a need to develop methods able to effectively combine weather information with genotype data to predict the performance of lines better. In this work, we develop a novel three-stage classifier to predict multi-class traits by combining three data types-genomic, weather, and secondary trait. The method addressed various challenges in this problem, such as confounding, differing sizes of data types, and threshold optimization. The method was examined in different settings, including binary and multi-class responses, various penalization schemes, and class balances. Then, our method was compared to standard machine learning methods such as random forests and support vector machines using various classification accuracy metrics and using model size to evaluate the sparsity of the model. The results showed that our method performed similarly to or better than machine learning methods across various settings. More importantly, the classifiers obtained were highly sparse, allowing for a straightforward interpretation of relationships between the response and the selected predictors.

Machine learning bridges omics sciences and plant breeding

Some of the biological knowledge obtained from fundamental research will be implemented in applied plant breeding. To bridge basic research and breeding practice, machine learning (ML) holds great promise to translate biological knowledge and omics data into precision-designed plant breeding. Here, we review ML for multi-omics analysis in plants, including data dimensionality reduction, inference of gene-regulation networks, and gene discovery and prioritization. These applications will facilitate understanding trait regulation mechanisms and identifying target genes potentially applicable to knowledge-driven molecular design breeding. We also highlight applications of deep learning in plant phenomics and ML in genomic selection-assisted breeding, such as various ML algorithms that model the correlations among genotypes (genes), phenotypes (traits), and environments, to ultimately achieve data-driven genomic design breeding.