Artificial intelligence (AI)-based protein modeling for the interpretation of grapevine genetic variants

The Institut Français De La Vigne Et Du Vin (IFV) is conducting many experiments on innovative winegrowing practices, which are emerging in companies in the sector, or which are still at the R&D stage for agricultural suppliers. The purpose of these practices may be to reduce environmental impact, to adapt vineyards to climate change, or to achieve other technical, economic or social aims. Whatever the objective, it is necessary to verify the relevance of these new practices, and in particular their environmental relevance, i.e. That at the very least, the changes in practices do not increase the environmental impact of the technical itineraries.

Wine grape berries respond to postharvest treatments with specific gaseous elicitors in terms of metabolic changes and composition. Short-term (3 days) high (30 KPa) CO2 treatment affects phenol compound concentration in skins of ‘Trebbiano toscano’ berries.

After Champagne popping, the first consumer’s observation is the shape of the cork stopper. Consumers expect a “mushroom shape”. Nevertheless, we sometimes observe a “barrel” shape due to inappropriate cork’s elastic properties. The aim of this study was to follow the loss of cork stopper resiliency during 26 months according to the density (d) of the cork in contact with the wine. 1680 disks were weighed + measured and divided in 6 density classes: High (H1 d= 0,19 g/cm3 – H2 d= 0,21 g/cm3), Medium (M, not studied) and Low (L1 d= 0,13 g/cm3 – L2 d= 0,14 g/cm3). Then, 138 technical cork stoppers were produced for each of the 4 studied groups. These corks consisted of an agglomerated natural cork granule body to which two natural cork disks were glued. A total of 552 bottles of sparkling wine were closed with these corks and open after 13, 19 and 26 months to follow cork resiliencies. Wine bottles were stored horizontally; thus, the external natural cork disks were in contact to the wine. During the 26 months of the study, highly significant differences (ANOVA) were observed between the resiliencies of H-corks and those of L-corks, whatever the time studied. The diameters of the L-corks were statistically higher than those of the H-corks. No significant differences were observed between L1 and L2 corks. At the opposite, differences were noted between H1 and H2 at 19 and 26 months. This could be explained by the heterogeneity of the resiliency that was higher for H-corks than for L-corks. Finally, the corks were visually (12 judges) divided in 3 classes corresponding to high (expected mushroom shape, i.e high resiliency), medium (irregular shape of the disk in contact with the wine and/or low premature deterioration of the expected resiliency) and low qualities (barrel shape = premature deterioration of the resiliency). The corks were also divided in 3 categories corresponding to 0-33%, 34-66% and 67-100% resiliency. A strong correlation was noted between the visual and the instrumental categorizations. This study strongly evidenced 1) the importance of the cork density on the cork stopper behaviour when opening the bottle and 2) the interest of an instrumental approach reflecting the consumer’s perception.

Temperature control is crucial in wine production, starting from grape harvest to the bottled wine storage. Climate change and global warming affect the timing of grape ripening, and harvesting is often done during hot summer days, influencing berry integrity, secondary metabolites potential, enzyme and oxidation phenomena, and even fermentation kinetics. To curb this phenomenon, pre-fermentative cold storage can help preserve the grapes and possibly increase the concentration of key secondary metabolites. In this study, the effect of grape pre-fermentative cold storage was assessed on the ‘Moscato bianco’ white grape cultivar, known for its varietal terpenes (65% of free terpenes represented by linalool and its derivatives) and widely used in Piedmont (Italy) to produce Asti DOCG wines.

A large varietal collection including over 1700 varieties was maintained in Conegliano, ITA, since the 1950s. Phenological data on a subset of 400 grape varieties including wine grapes, table grapes, and raisins were acquired at bud break, flowering, veraison, and ripening since 1964. Despite the efforts in maintaining and acquiring data over such an extensive collection, the data set has varying degrees of missing cases depending on the variety and the year. This is ubiquitous in phenology datasets with significant size and length. In this work, we evaluated four state-of-the-art methods to estimate missing values in this phenological series: k-Nearest Neighbour (kNN), Multivariate Imputation by Chained Equations (mice), MissForest, and Bidirectional Recurrent Imputation for Time Series (BRITS). For each phenological stage, we evaluated the performance of the methods in two ways. 1) On the full dataset, we randomly hold-out 10% of the true values for use as a test set and repeated the process 1000 times (Monte Carlo cross-validation). 2) On a reduced and almost complete subset of varieties, we varied the percentage of missing values from 10% to 70% by random deletion. In all cases, we evaluated the performance on the original values using normalized root mean squared error. For the full dataset we also obtained performance statistics by variety and by year. MissForest provided average errors of 17% (3 days) at budbreak, 14% (4 days) at flowering, 14.5% (7 days) at veraison, and 17% (3 days) at maturity. We completed the imputations of the Conegliano dataset, one of the world’s most extensive and varied phenological time series and a steppingstone for future climate change studies in grapes. The dataset is now ready for further analysis, and a rigorous evaluation of imputation errors is included.