No abstract

Genomic prediction tools support crop breeding based on statistical methods, such as the genomic best linear unbiased prediction (GBLUP). However, these tools are not designed to capture non-linear relationships within multi-dimensional datasets, or deal with high dimension datasets such as imagery collected by unmanned aerial vehicles. Machine learning (ML) algorithms have the potential to surpass the prediction accuracy of current tools used for genotype to phenotype prediction, due to their capacity to autonomously extract data features and represent their relationships at multiple levels of abstraction. This review addresses the challenges of applying statistical and machine learning methods for predicting phenotypic traits based on genetic markers, environment data, and imagery for crop breeding. We present the advantages and disadvantages of explainable model structures, discuss the potential of machine learning models for genotype to phenotype prediction in crop breeding, and the challenges, including the scarcity of high-quality datasets, inconsistent metadata annotation and the requirements of ML models.

Genomic selection approaches have increased the speed of plant breeding, leading to growing crop yields over the last decade. However, climate change is impacting current and future yields, resulting in the need to further accelerate breeding efforts to cope with these changing conditions. Here we present approaches to accelerate plant breeding and incorporate nonadditive effects in genomic selection by applying state-of-the-art machine learning approaches. These approaches are made more powerful by the inclusion of pangenomes, which represent the entire genome content of a species. Understanding the strengths and limitations of machine learning methods, compared with more traditional genomic selection efforts, is paramount to the successful application of these methods in crop breeding. We describe examples of genomic selection and pangenome-based approaches in crop breeding, discuss machine learning-specific challenges, and highlight the potential for the application of machine learning in genomic selection. We believe that careful implementation of machine learning approaches will support crop improvement to help counter the adverse outcomes of climate change on crop production.

Assessing crop production in the field often requires breeders to wait until the end of the season to collect yield-related measurements, limiting the pace of the breeding cycle. Early prediction of crop performance can reduce this constraint by allowing breeders more time to focus on the highest-performing varieties. Here, we present a multimodal deep learning model for predicting the performance of maize (Zea mays) at an early developmental stage, offering the potential to accelerate crop breeding. We employed multispectral images and eight vegetation indices, collected by an uncrewed aerial vehicle approximately 60 days after sowing, over three consecutive growing cycles (2017, 2018 and 2019). The multimodal deep learning approach was used to integrate field management and genotype information with the multispectral data, providing context to the conditions that the plants experienced during the trial. Model performance was assessed using holdout data, in which the model accurately predicted the yield (RMSE 1.07 t/ha, a relative RMSE of 7.60% of 16 t/ha, and R2 score 0.73) and identified the majority of high-yielding varieties, outperforming previously published models for early yield prediction. The inclusion of vegetation indices was important for model performance, with a normalized difference vegetation index and green with normalized difference vegetation index contributing the most to model performance. The model provides a decision support tool, identifying promising lines early in the field trial.

High throughput phenotyping (HTP) platforms are capable of monitoring the phenotypic variation of plants through multiple types of sensors such as RGB cameras, hyperspectral sensors and computed tomography, which can be associated with environmental and genotypic data. Because of the wide range of information provided, HTP datasets represent a valuable asset to characterise crop phenotypes. As HTP becomes widely employed with more tools and data being released, it is important that researchers are aware of these resources and how they can be applied to accelerate crop improvement. Researchers may exploit these datasets either for phenotype comparison or employ them as a benchmark to assess tool performance and to support the development of tools that are better at generalising between different crops and environments. In this review, we describe the use of image-based HTP for yield prediction, root phenotyping, development of climate-resilient crops, detecting pathogen and pest infestation and quantitative trait measurement. We emphasise the need for researchers to share phenotypic data, and offer a comprehensive list of available datasets to assist crop breeders and tool developers to leverage these resources in order to accelerate crop breeding.

Pangenomes are a rich resource to examine the genomic variation observed within a species or genera, supporting population genetics studies, with applications for the improvement of crop traits. Major crop species such as maize (Zea mays), rice (Oryza sativa), Brassica (Brassica spp.), and soybean (Glycine max) have had pangenomes constructed and released, and this has led to the discovery of valuable genes associated with disease resistance and yield components. However, pangenome data are not available for many less prominent crop species that are currently under-utilised. Despite many under-utilised species being important food sources in regional populations, the scarcity of genomic data for these species hinders their improvement. Here, we assess several under-utilised crops and review the pangenome approaches that could be used to build resources for their improvement. Many of these under-utilised crops are cultivated in arid or semi-arid environments, suggesting that novel genes related to drought tolerance may be identified and used for introgression into related major crop species. In addition, we discuss how previously collected data could be used to enrich pangenome functional analysis in genome-wide association studies (GWAS) based on studies in major crops. Considering the technological advances in genome sequencing, pangenome references for under-utilised species are becoming more obtainable, offering the opportunity to identify novel genes related to agro-morphological traits in these species.

Pangenomes aim to represent the complete repertoire of the genome diversity present within a species or cohort of species, capturing the genomic structural variance between individuals. This genomic information coupled with phenotypic data can be applied to identify genes and alleles involved with abiotic stress tolerance, disease resistance, and other desirable traits. The characterisation of novel structural variants from pangenomes can support genome editing approaches such as Clustered Regularly Interspaced Short Palindromic Repeats and CRISPR associated protein Cas (CRISPR-Cas), providing functional information on gene sequences and new target sites in variant-specific genes with increased efficiency. This review discusses the application of pangenomes in genome editing and crop improvement, focusing on the potential of pangenomes to accurately identify target genes for CRISPR-Cas editing of plant genomes while avoiding adverse off-target effects. We consider the limitations of applying CRISPR-Cas editing with pangenome references and potential solutions to overcome these limitations.

Soybean (Glycine max) is a legume species of significant economic and nutritional value. The yield of soybean continues to increase with the breeding of improved varieties, and this is likely to continue with the application of advanced genetic and genomic approaches for breeding. Genome technologies continue to advance rapidly, with an increasing number of high-quality genome assemblies becoming available. With accumulating data from marker arrays and whole-genome resequencing, studying variations between individuals and populations is becoming increasingly accessible. Furthermore, the recent development of soybean pangenomes has highlighted the significant structural variation between individuals, together with knowledge of what has been selected for or lost during domestication and breeding, information that can be applied for the breeding of improved cultivars. Because of this, resources such as genome assemblies, SNP datasets, pangenomes and associated databases are becoming increasingly important for research underlying soybean crop improvement.