Vine Bioinformatics

Principal Investigator: Prof. Hans Binder

University: University of Leipzig

Research Group: Kristina Margaryan, Maria Nikoghosyan, Tomas Konecny, Anush Baloyan, Hripsime Gasoyan, Emma Hovhannisyan, Levon Galstyan, Duan Shengchang

Duration: 2023-2027

Co-implementing and hosting partner: Armenian Bioinformatics Institute (ABI) and Institute of Molecular Biology

Project Importance The cultivated grape (Vitis vinifera) has become the world’s leading fruit crop. Grape is unique not only because it is a major global perennial crop but also for its historical and cultural connections with the development of humans. One of the main challenges for viticulture is to sustainably maintain the production of high-quality grape varieties in the face of climate change. Current models predict an increasing disease pressure for grapes, mainly because of warmer and partly drier conditions. Several knowledge gaps exist regarding these challenges.

1. Understanding the whole genome diversity of vine: With the advent of next generation sequencing, whole genome data become increasingly available. This data is permanently growing: genomic datasets of hundreds of vine accessions are produced and added each year. The genetic information hidden in this data which in size is roughly comparable with that of the human genome is by far not extracted or even understood.

2. Functional Genomics: The genomic information becomes functional after transcription and translation. Omics-data (e.g., transcriptomics, epigenetics, proteomics) are increasingly generated for vine and other cultivated plants to study functional genomics with impact for improving yield, quality, and resistance. However, many issues, related to, e.g., development, stress-response, and resistance are still understudied. This opens novel opportunities to close this gap by analyzing such datasets.

3. Machine learning of multidimensional plant omics data. The flood of omics data generated require appropriate processing and analysis for their transformation into useful knowledge. Bioinformatics methods including algorithm development and data science techniques are needed to accomplish these tasks. Machine learning and artificial intelligence in combination with “domain knowledge” are the most promising approaches for extracting the information hidden in these highly complex and multidimensional data.

4. Translation into FAIR viticulture. The results of basic research must be translated into practice. As a first step, the generated information must be made available to the scientific community under the FAIR (Findable, Accessible, Interoperable, Re-usable) standards for scientific data. This requirement applies to primary sequencing data, secondary genomic, as well as phenotypic characteristics of the accessions, and also the results of the downstream analyses together with the methods and the models used. Presently, this information is either not available or it is spread over many disjunct data repositories hampering their effective exploitation.

Expected Results and Impact

1. We will decipher the genomic history and diversity of Armenian vine accessions by performing a detailed analysis of Armenian vine genomic data. We will study the primary origin of vine cultivation, its dissemination in space and time with a special focus on Armenia, genetic relations between wild and cultivated vine, as well as genetic resistance mechanisms and genes acting against stress factors (“R-genes”). We will refine the “cross road” region of the genetic relatedness of vine accessions located in the South Caucasus which links different vine-dissemination routes.

2. We will apply functional genomics by analyzing transcriptomic and epigenetic mechanisms of genomic regulation. We will achieve a deeper understanding of these mechanisms with potential impact for developing biotechnological measures to overcome stress related problems of vine cultivation, e.g., by identifying key-genes and/or pathways as candidates for interventions.

3. The methods part of the project is of paramount importance for accomplishing the tasks of the other objectives. The “plant-oposSOM” pipeline - we will set standards for the analysis of omics data of plants. This ML-based method will allow us to deconvolute the multi-dimensional omics-state space of plant genomes, perform its functional annotation and accomplish additional subtasks such as selecting conditions, genes and mechanisms of interest, portrayal of each particular omics-state, all available via an extensive reporting and visualization platform. Plant-oposSOM will enable the effective analysis of novel data collected and addressed in this project and also in future studies.

4. The Genomic Atlas of Vine diversity in Armenia will collect results obtained in the frame of points 1-3. It will make them visible and available to the international research community. The Atlas will provide web-services, e.g., a frontend for browsing genetic and other omics data interactively. The Atlas will not only provide data online, but also analyses and results as well as methods and models developed and used for analysis. The Atlas will become the starting nucleus for the Armenian Life Bank, which is planned at ABI to become the central repository of genomic data collected in Armenia.