Genomic Pattern Recognition in DUS testing for Barley
To be marketed and/or awarded Plant Breeders Rights (PBR) in the UK, all agricultural varieties must pass DUS (Distinctness, Uniformity and Stability) testing. This ensures that new varieties are unique, with distinctness determined by visually comparing candidate varieties against other varieties in common knowledge (the ‘variety collection’).
In this Defra-funded project we are exploring ways to accelerate variety registration using genomic prediction approaches. Working on barley, we are refining and optimising our machine learning prediction models, focussed on prediction of individual barley DUS characteristics (phenotypes), to facilitate earlier selection of similar varieties from the variety collection for field distinctness assessments. Predictions models are being tested in parallel to current DUS testing procedures, with consideration of logistical and technological challenges for future implementation. Software for user-implementation of the finalised models is also being developed to support the use of the analysis pipeline by DUS testing centres.
Duration
September 2025-March 2028
Funder













