Combining Vitis OneGene causality approach with phylogenetic and cistrome analyses to study the laccase, dirigent and peroxidase gene networks

The availability of transcriptomic data organized in coherent databases and the development of causal inference methods have paved the way to gene network analysis in grapevine. The identification and visualization of gene-to-gene interactions as networks can contribute to several aims, such as the identification of hub genes or modules, the prediction of gene function, the integration of different -omics data and the comparison of the networks found across species. OneGenE, an in house developed causality-based data mining tool, has been applied to the Vitis transcriptomic data set VESPUCCI to produce a list of directly interacting genes, for each gene of the grapevine genome. Due to the computational power requirement of the OneGenE algorithm, it has been run as part of the gene@home project, a distributed computing project relying on thousands of volunteers computers by means of the TN-Grid, an infrastructure based on BOINC system. Finally, bioinformatic tools have been built in house to reconstruct association gene networks using the OneGenE generated lists. All this information and analysis tools are now available to the community through the Vitis OneGenE website. Here, we analyse the OneGenE association networks of three gene families putatively involved in lignin biosynthesis and integrate this information with phylogenetic analyses and DNA affinity purification sequencing (DAP-seq) data. The enzymes laccase, peroxidase and Dirigent proteins, present in the grapevine genome as large gene families, catalyse the final steps of monolignol polymerization in lignin biosynthesis. However, our results strongly suggest the evolutionary divergence of a group of them possibly involved in the oligomerization of resveratrol to produce oligomeric stilbenoids, phytoalexins with antifungal properties. Although experimental validation is undoubtedly required to confirm these findings, they represent an interesting example of how bioinformatic analyses based on the reuse of existing data can support and accelerate knowledge discovery in grapevine.