Using machine learning to increase genetic gain in canola blackleg resistance breeding

This project will extend current research into deep learning and the genomics of disease resistance and apply findings to improve the blackleg resistance of Australian canola cultivars. Neural networks are a series of algorithms that mimic the operations of a human brain to recognise relationships between vast amounts of data. In this project, neural networks will be trained to associate canola blackleg resistance phenotypes with gene expression, whole genome sequence, and single-nucleotide polymorphism data from a range of different data sources. The neural networks will identify non-additive genetic effects of blackleg resistance that cannot be easily identified using standard multivariate statistical models.