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

Genomic prediction of seed nutritional traits in biparental families of oat (Avena sativa)

Selection for more nutritious crop plants is an important goal of plant breeding to improve food quality and contribute to human health outcomes. While there are efforts to integrate genomic prediction to accelerate breeding progress, an ongoing challenge is identifying strategies to improve accuracy when predicting within biparental populations in breeding programs. We tested multiple genomic prediction methods for 12 seed fatty acid content traits in oat (Avena sativa L.), as unsaturated fatty acids are a key nutritional trait in oat. Using two well-characterized oat germplasm panels and other biparental families as training populations, we predicted family mean and individual values within families. Genomic prediction of family mean exceeded a mean accuracy of 0.40 and 0.80 using an unrelated and related germplasm panel, respectively, where the related germplasm panel outperformed prediction based on phenotypic means (0.54). Within family prediction accuracy was more variable: training on the related germplasm had higher accuracy than the unrelated panel (0.14-0.16 and 0.05-0.07, respectively), but variability between families was not easily predicted by parent relatedness, segregation of a locus detected by a genome-wide association study in the panel, or other characteristics. When using other families as training populations, prediction accuracies were comparable to the related germplasm panel (0.11-0.23), and families that had half-sib families in the training set had higher prediction accuracy than those that did not. Overall, this work provides an example of genomic prediction of family means and within biparental families for an important nutritional trait and suggests that using related germplasm panels as training populations can be effective.

Leveraging prior biological knowledge improves prediction of tocochromanols in maize grain

With an essential role in human health, tocochromanols are mostly obtained by consuming seed oils; however, the vitamin E content of the most abundant tocochromanols in maize (Zea mays L.) grain is low. Several large-effect genes with cis-acting variants affecting messenger RNA (mRNA) expression are mostly responsible for tocochromanol variation in maize grain, with other relevant associated quantitative trait loci (QTL) yet to be fully resolved. Leveraging existing genomic and transcriptomic information for maize inbreds could improve prediction when selecting for higher vitamin E content. Here, we first evaluated a multikernel genomic best linear unbiased prediction (MK-GBLUP) approach for modeling known QTL in the prediction of nine tocochromanol grain phenotypes (12-21 QTL per trait) within and between two panels of 1,462 and 242 maize inbred lines. On average, MK-GBLUP models improved predictive abilities by 7.0-13.6% when compared with GBLUP. In a second approach with a subset of 545 lines from the larger panel, the highest average improvement in predictive ability relative to GBLUP was achieved with a multi-trait GBLUP model (15.4%) that had a tocochromanol phenotype and transcript abundances in developing grain for a few large-effect candidate causal genes (1-3 genes per trait) as multiple response variables. Taken together, our study illustrates the enhancement of prediction models when informed by existing biological knowledge pertaining to QTL and candidate causal genes.

Generalizable approaches for genomic prediction of metabolites in plants

Plant metabolites are important traits for plant breeders seeking to improve nutrition and agronomic performance yet integrating selection for metabolomic traits can be limited by phenotyping expense and degree of genetic characterization, especially of uncommon metabolites. As such, developing generalizable genomic selection methods based on biochemical pathway biology for metabolites that are transferable across plant populations would benefit plant breeding programs. We tested genomic prediction accuracy for >600 metabolites measured by gas chromatography-mass spectrometry (GC-MS) and liquid chromatography-mass spectrometry (LC-MS) in oat (Avena sativa L.) seed. Using a discovery germplasm panel, we conducted metabolite genome-wide association study (mGWAS) and selected loci to use in multikernel models that encompassed metabolome-wide mGWAS results or mGWAS from specific metabolite structures or biosynthetic pathways. Metabolite kernels developed from LC-MS metabolites in the discovery panel improved prediction accuracy of LC-MS metabolite traits in the validation panel consisting of more advanced breeding lines. No approach, however, improved prediction accuracy for GC-MS metabolites. We ranked model performance by metabolite and found that metabolites with similar polarity had consistent rankings of models. Overall, testing biological rationales for developing kernels for genomic prediction across populations contributes to developing frameworks for plant breeding for metabolite traits.

Recent applications of metabolomics in plant breeding

Metabolites play a central role in maintaining organismal life and in defining crop phenotypes, such as nutritional value, fragrance, color, and stress resistance. Among the 'omes' in biology, the metabolome is the closest to the phenotype. Consequently, metabolomics has been applied to crop improvement as method for monitoring changes in chemical compositions, clarifying the mechanisms underlying cellular functions, discovering markers and diagnostics, and phenotyping for mQTL, mGWAS, and metabolite-genome predictions. In this review, 359 reports of the most recent applications of metabolomics to plant breeding-related studies were examined. In addition to the major crops, more than 160 other crops including rare medicinal plants were considered. One bottleneck associated with using metabolomics is the wide array of instruments that are used to obtain data and the ambiguity associated with metabolite identification and quantification. To further the application of metabolomics to plant breeding, the features and perspectives of the technology are discussed.

Multi-omics prediction of oat agronomic and seed nutritional traits across environments and in distantly related populations

Integration of multi-omics data improved prediction accuracies of oat agronomic and seed nutritional traits in multi-environment trials and distantly related populations in addition to the single-environment prediction. Multi-omics prediction has been shown to be superior to genomic prediction with genome-wide DNA-based genetic markers (G) for predicting phenotypes. However, most of the existing studies were based on historical datasets from one environment; therefore, they were unable to evaluate the efficiency of multi-omics prediction in multi-environment trials and distantly related populations. To fill those gaps, we designed a systematic experiment to collect omics data and evaluate 17 traits in two oat breeding populations planted in single and multiple environments. In the single-environment trial, transcriptomic BLUP (T), metabolomic BLUP (M), G + T, G + M, and G + T + M models showed greater prediction accuracy than GBLUP for 5, 10, 11, 17, and 17 traits, respectively, and metabolites generally performed better than transcripts when combined with SNPs. In the multi-environment trial, multi-trait models with omics data outperformed both counterpart multi-trait GBLUP models and single-environment omics models, and the highest prediction accuracy was achieved when modeling genetic covariance as an unstructured covariance model. We also demonstrated that omics data can be used to prioritize loci from one population with omics data to improve genomic prediction in a distantly related population using a two-kernel linear model that accommodated both likely casual loci with large-effect and loci that explain little or no phenotypic variance. We propose that the two-kernel linear model is superior to most genomic prediction models that assume each variant is equally likely to affect the trait and can be used to improve prediction accuracy for any trait with prior knowledge of genetic architecture.

Multi-omics prediction of oat agronomic and seed nutritional traits across environments and in distantly related populations

Integration of multi-omics data improved prediction accuracies of oat agronomic and seed nutritional traits in multi-environment trials and distantly related populations in addition to the single-environment prediction. Multi-omics prediction has been shown to be superior to genomic prediction with genome-wide DNA-based genetic markers (G) for predicting phenotypes. However, most of the existing studies were based on historical datasets from one environment; therefore, they were unable to evaluate the efficiency of multi-omics prediction in multi-environment trials and distantly related populations. To fill those gaps, we designed a systematic experiment to collect omics data and evaluate 17 traits in two oat breeding populations planted in single and multiple environments. In the single-environment trial, transcriptomic BLUP (T), metabolomic BLUP (M), G + T, G + M, and G + T + M models showed greater prediction accuracy than GBLUP for 5, 10, 11, 17, and 17 traits, respectively, and metabolites generally performed better than transcripts when combined with SNPs. In the multi-environment trial, multi-trait models with omics data outperformed both counterpart multi-trait GBLUP models and single-environment omics models, and the highest prediction accuracy was achieved when modeling genetic covariance as an unstructured covariance model. We also demonstrated that omics data can be used to prioritize loci from one population with omics data to improve genomic prediction in a distantly related population using a two-kernel linear model that accommodated both likely casual loci with large-effect and loci that explain little or no phenotypic variance. We propose that the two-kernel linear model is superior to most genomic prediction models that assume each variant is equally likely to affect the trait and can be used to improve prediction accuracy for any trait with prior knowledge of genetic architecture.

Improving Genomic Prediction for Seed Quality Traits in Oat (Avena sativa L.) Using Trait-Specific Relationship Matrices

The observable phenotype is the manifestation of information that is passed along different organization levels (transcriptional, translational, and metabolic) of a biological system. The widespread use of various omic technologies (RNA-sequencing, metabolomics, etc.) has provided plant genetics and breeders with a wealth of information on pertinent intermediate molecular processes that may help explain variation in conventional traits such as yield, seed quality, and fitness, among others. A major challenge is effectively using these data to help predict the genetic merit of new, unobserved individuals for conventional agronomic traits. Trait-specific genomic relationship matrices (TGRMs) model the relationships between individuals using genome-wide markers (SNPs) and place greater emphasis on markers that most relevant to the trait compared to conventional genomic relationship matrices. Given that these approaches define relationships based on putative causal loci, it is expected that these approaches should improve predictions for related traits. In this study we evaluated the use of TGRMs to accommodate information on intermediate molecular phenotypes (referred to as endophenotypes) and to predict an agronomic trait, total lipid content, in oat seed. Nine fatty acids were quantified in a panel of 336 oat lines. Marker effects were estimated for each endophenotype, and were used to construct TGRMs. A multikernel TRGM model (MK-TRGM-BLUP) was used to predict total seed lipid content in an independent panel of 210 oat lines. The MK-TRGM-BLUP approach significantly improved predictions for total lipid content when compared to a conventional genomic BLUP (gBLUP) approach. Given that the MK-TGRM-BLUP approach leverages information on the nine fatty acids to predict genetic values for total lipid content in unobserved individuals, we compared the MK-TGRM-BLUP approach to a multi-trait gBLUP (MT-gBLUP) approach that jointly fits phenotypes for fatty acids and total lipid content. The MK-TGRM-BLUP approach significantly outperformed MT-gBLUP. Collectively, these results highlight the utility of using TGRM to accommodate information on endophenotypes and improve genomic prediction for a conventional agronomic trait.