The study performed a meta‑transcriptomic analysis of over twenty drought versus control experiments in Vitis vinifera and two hybrid rootstocks, identifying a core set of 4,617 drought‑responsive genes. Using transcription factor binding motif enrichment and random‑forest machine learning, gene regulatory networks were built, revealing key regulators such as ABF2, MYB30A, and a novel HMG‑box protein. These regulators and network hierarchies provide candidate targets for breeding and biotechnological improvement of grapevine drought tolerance.

The study examined molecular responses in grapevine leaves with and without esca symptoms, using metabolite profiling, RNA‑seq and whole‑genome bisulfite sequencing. Metabolic and transcriptomic changes were confined to symptomatic leaves and linked to local DNA‑methylation alterations, while asymptomatic leaves showed distinct but overlapping methylation patterns, some present before symptoms, indicating potential epigenetic biomarkers for early disease detection.

The study reanalyzed 1,107 public grapevine RNA‑seq datasets to build condition‑specific gene expression atlases and a whole‑genome co‑expression network associated with drought stress, and deployed these resources via a web‑based Hydric Stress Atlas App. Network topology analysis identified candidate hub genes that could serve as molecular markers or targets for gene editing to improve drought tolerance in Vitis vinifera.

The study evaluated high‑throughput spectroscopy and poro‑fluorometry to predict leaf morphological and ecophysiological traits in a grapevine diversity panel under well‑watered and drought conditions. Spectroscopy reliably estimated leaf mass per area and water content, while poro‑fluorometry accurately predicted net CO2 assimilation, and the derived predicted traits showed substantial broad‑sense heritability. These results demonstrate that non‑destructive, rapid phenotyping tools can support genetic analyses of drought‑related traits in grapevine.