Using genomic prediction with crop growth models enables the prediction of associated traits in wheat

Abstract

Crop growth models (CGM) can predict the performance of a cultivar in untested environments by sampling genotype-specific parameters (GSPs). As they cannot predict the performance of new cultivars, it has been proposed to integrate CGMs with whole genome prediction (WGP) to combine the benefits of both models. Here, we used a CGM-WGP model to predict the performance of new wheat genotypes that do not have phenotypic records in the reference population. The CGM was designed to predict phenology, nitrogen, and biomass traits. The CGM-WGP simulated more heritable GSPs compared to the CGM and gave smaller errors for the observed phenotypes. The WGP model performed better when predicting yield, grain number and grain protein content, but showed comparable performance to the CGM-WGP for heading and physiological maturity dates. However, the CGM-WGP model was able to predict unobserved traits (for which there were no phenotypic records in the reference population) through the prediction of biophysical relations between the GSPs and environmental inputs. CGM-WGP also showed superior performance when predicting unrelated individuals that clustered separately from the reference population in a PCA plot. Genome-wide association mapping linked the GSPs to previously reported loci for agronomic and physiological traits. Our results demonstrate new advantages for CGM-WGP modelling and suggest future efforts should focus on calibrating CGM-WGPs using high-throughput phenotypic measures that are cheaper and less laborious to collect.

Current Status and Future Prospects of Head Rice Yield

“…Future studies should investigate the effect of predicted phenology traits on the prediction accuracy of untested environments. Alternatively, untested or future climates can be predicted in a more straightforward CGM-WGP, which integrates crop growth models and genomic prediction [196].…”

Using machine learning to integrate genetic and environmental data to model genotype-by-environment interactions

“…However, we use it in a fundamentally different way as we utilize the historic yield from hybrids that are not necessarily related to the target hybrids to characterize each environment. ECs are another type of data utilized in our model that is becoming more common in GP models (Bustos-Korts et al 2019; Onogi 2022; Jighly et al 2022). Our results confirmed that ECs such as the ones derived from the APSIM crop model are useful features in improving prediction abilities.…”

Crop modeling suggests limited transpiration would increase yield of sorghum across drought-prone regions of the United States

“…Specifically, this study assessed management practices to leverage the performance of sorghum hybrids ( Figure 5B ), envirotyped the target population of environments with distinctive water stress patterns ( Figures 5A, B ), and quantified the impact of the LT trait in the production sorghum area. Crop model applications to support breeding exist for different crops and diverse geographies ( Technow et al., 2015 ; Chenu et al., 2017 ; Hammer et al., 2019b ; Jighly et al., 2023 ). For instance, a study deployed a rice crop model to hypothesize the lack of effectiveness in breeding drought-tolerant upland rice in Brazil ( Heinemann et al., 2019 ), and other studies improved phenomic prediction by integrating crop models and genomic prediction ( Cooper et al., 2002 ; Heslot et al., 2014 ; Crossa et al., 2022 ).…”