Integration of Genomics and Crop Modeling for Prediction of Complex Traits.

Marker-assisted selection has been an integral component of rice breeding for over two decades. However, its use has largely been restricted to traits controlled by major genes, and has been less useful for complex traits. Genome-wide association mapping (GWAS) and genomic selection (GS) are promising technologies in trait identification and breeding design. However, neither approach considers the underlying physiological processes of plant growth and development, and as such, has limited ability to separate G×E interactions and to integrate the contributions of multiple component traits.

Process-based crop modeling applies a methodology involving trait and process decomposition, and component integration. It offers the ability to separate G×E interactions and integrate the contributions of multiple component traits and is being increasingly realized as a promising technology for use in advancing genomics research.

We first provide a brief review on GWAS and genomic prediction and recent approaches on integrating crop models with genomic technology and their limitations and then present a framework for complex trait decomposition and component trait integration. We discuss the limitation of existing models and the need for change in model structure to provide the capability to drill down the biological scale and narrow the gap between component traits and gene actions.

This research presents a potential path to integrate genomics with crop modeling to accelerate crop breeding and further advance remarkable progress in genomic research and to bridge the gaps between genomics and complex trait prediction.