Streamlining Phenotyping for the Energy Crop Miscanthus

"Breeding a new crop as a bioenergy feedstock requires the application of a wide range of techniques including genetics, biochemistry, physiology, bioinformatics and systems biology modelling. The achievement of sustainability is a multiple-optimisation process that must take account of many different factors. Over the last few years IBERS has established experience and knowledge of using computational and modelling techniques to inform breeding. This has been achieved through BBSRC funded Sustainable Bioenergy Centre (BSBEC) (BBG0162161), and an Industrial CASE studentship (BBH016481) to build predictive crop models for Miscanthus that associate the performance and traits of particular genotypes in various environments. The proposed project will build on this experience of using existing crop models such as MiscanMod for establishing computational models to identify what is the minimum measurement required for phenotyping in order to achieve the optimal effect either for phenotypic selection (PS) and/or genomic selection (GS). This will enable a dramatic reduction in phenotyping costs while retaining the same precision in selection for breeding.
This project will use data-mining, machine-learning and modelling techniques that are well suited to handling the multidimensional structure of the data, which contains a large numbers of descriptors (many are best represented relationally) in order to create association models. The typical aspects of Miscanthus data that lend themselves to relational description are: temporal relations; spatial relations across fields and plants; pedigree relationships between genotypes, etc. A combination of methodologies will be used for this research such as neural networks, Bayesian regression, decision trees and random forests etc. To evaluate and compare the performance of these approaches with standard statistical genetic methods, we will use standard re-sampling approaches, including cross-validation and bootstrapping.
This project is fully aligned with the strategic priorities of the BBSRC and its commitment of pathway to impact in the area of bioenergy and data-driven biology. This project will create a great impact toward the advancement in science by improving the breeding process and by addressing the global challenge of delivering a sustainable future.
The primary end beneficiaries of this work will be the Defra/BBSRC funded Miscanthus breeding GIANT Link programme (in which Ceres is the industrial partner), the BBSRC Institute Strategic Programme Grant (ISPG - 'Energy Grass and Biorefinery'), the IBERS/Ceres multi-location trials in the US and Europe and the newly established genome wide association mapping (GWAS) on Miscanthus at IBERS. However the new tools would equally be applicable to a wide range of organisms and the newly created predictive models would also serve as a reference for use on other crops to achieve effective GS. It is expected that the improvements in the phenotying of Miscanthus will bring benefits throughout the fuel chain from producer through to consumer in terms of a sustainable rural economy, food and fuel security and environmental benefits. Collaboration between IBERS and Ceres on Miscanthus enables the translation of model and crop plant research into improved Miscanthus varieties and is an excellent example of public-private transatlantic collaboration to combat global climate change."