Genetic and molecular determinism of sorghum grain protein content and digestibility. Towards an optimisation of prediction capacities

Despite its agronomic advantages, the low digestibility of grain sorghum storage proteins by gastrointestinal proteases is a major obstacle to the wider use of sorghum for human and animal nutrition. Genetic variability exists for this trait, which suggests the possibility of developing better adapted varieties. A better understanding of the molecular mechanisms involved in protein accumulation and digestibility is needed in addition to the development of more efficient breeding methods to achieve this goal. These two aspects constitute the objectives of this thesis project.

For that purpose, a coupled genomic and genetic approach will be used:

The gene network analysis data already available will be exploited to identify new transcription factors potentially regulating the mechanisms of protein pooling, and their role will be tested using a simplified overexpression system.

Genotyping and phenotyping data (using NIRS, i.e. Near Infra Red Spectroscopy) will be produced as part of the CASDAR Nitrosorg programme on which this thesis is based using a sample representing world diversity. This data will be aggregated to analyse the genetic determinism of quality traits based on an association genetics approach. The data obtained will also be used to develop genomic prediction models. Models will be built using spectral data (NIRS) alone (phenomenal selection) or in combination with molecular information (phenomenal selection coupled with major effect zones of the genome and genomic prediction approaches) to predict the performance of varieties.