The dynamic berry transcriptome during postharvest dehydration. A, PCA score scatter plot of the model obtained for transcripts detected by microarray analysis during postharvest dehydration in each variety (explained variance = 44.7%). Samples, corresponding to the averaged value of the three biological replicates at each stage, are colored according to variety. CO, Corvina; SA, Sangiovese; ME, Merlot; SY, Syrah; OS, Oseleta; CA, Cabernet Sauvignon. B, PCA correlation loading plot of the PCA model. Groups of genes more correlated to the first and second components are shown in different colors: first principal component negatively correlated, yellow; first principal component positively correlated, light blue; second principal component negatively correlated, green; second principal component positively correlated, red. The most represented functional categories for each group of genes are indicated by the arrows.