Bringing the voice of consumers into plant breeding with Bayesian modelling

Improving flavour quality traits in fruit breeding calls for innovative consumer-oriented product development. This paper explores the potential of marker-assisted breeding from genomics and consumer-based quality-improvement models from marketing, and exploits the progresses at both sides as technology push and market pull. An integrative and cross-disciplinary quality-improvement model is proposed based on Bayesian modelling. This Bayesian modelling allows for the integration of elicited knowledge of breeders and flavour researchers concerning the degree of causal associations of metabolites and flavour quality traits of fruits in the model. We also present the flavour quality improvement challenge as a multi-criteria optimization process and show the potential and current limitations of the proposed model. Insights gained from the model would help flavour researchers determine the optimum concentration of flavour-affecting metabolites which could be used for further DNA marker development. These ideas and concepts will help translate consumer-desired product features into genomic information, ultimately resulting in successful new cultivars