
CONVOLUTIONAL NEURAL NETWORK TO PREDICT GENETIC GROUP AND SULFUR TOLERANCE OF BRETTANOMYCES BRUXELLENSIS
Abstract
The spoilage yeast Brettanomyces bruxellensis presents many strain dependent characteristics, particularly sulfur tolerance (1). Climate change and the evolution of oenological practices are at the origin of matrices with low levels of sulfur dioxide and higher pH. These parameters favor the development of this yeast and lead to serious financial losses for winemakers (2). Thus, it is essential to discriminate B. bruxellensis isolates at the strain level in order to predict their stress resistance capacities. Few predictive tools are available to reveal intraspecific diversity within B. bruxellensis species; also, they require expertise and can be expensive. In this study, to make analysis even faster, we further investigated the correlation between genetic groups previously described (3) and cell polymorphism using the analysis of optical microscopy images via deep learning. A Convolutional Neural Network (CNN) was trained and allowed the discrimination of B. bruxellensis isolates in 4 of the 6 genetic groups (GG), with an accuracy of 96.6% (4). Future works will have to be done for the no tested genetic groups. But already these results confirm the possibility to develop a tool allowing to determine the tolerance of a contaminant, in a short time, in order to help wine industry professionals to choose the appropriate corrective measure.
DOI:
Issue: OENO Macrowine 2023
Type: Poster
Authors
Contact the author*
Keywords
Brettanomyces bruxellensis, deep learning, cell morphology, genetic groups