Our research group specializes in transforming data from the DSMZ Digital Diversity platform into RDF (Resource Description Framework) format, enabling structured and machine-readable data. This data is made available via SPARQL endpoints, allowing for flexible and efficient querying by researchers and external systems. We also develop and maintain the DSMZ Digital Diversity Ontology (D3O), which serves as a uniform framework for organizing and linking the diverse datasets. This ontology helps standardize the way biological data is represented and ensures consistency across various domains.

Many of our databases, such as BRENDA with over 130,000 manually annotated references, depend heavily on the manual annotation of primary literature. This process can be time-consuming and labor-intensive. To accelerate it, we are developing a novel annotation pipeline that incorporates natural language processing (NLP) techniques alongside intuitive user interfaces. These models assist human annotators by streamlining the annotation process, allowing for faster and more efficient curation of scientific literature.

We utilize advanced AI-driven methods to extract novel insights from our existing datasets, enabling the discovery of previously unknown relationships and knowledge. These predictive techniques enhance our ability to generate meaningful conclusions from large-scale biological data.

In addition, we are working on making this knowledge more accessible by integrating AI-driven tools for documentation and user assistance. These tools will streamline the process of understanding and interacting with complex datasets, making it easier for researchers and other users to navigate, interpret, and apply the information effectively.

We are actively developing novel databases and applications to leverage our full range of digital resources. This includes a new plant virus database, a toolbox that uses natural language processing (NLP) to automatically assign ontology terms to isolation sources, and graph visualizations for exploring knowledge graphs. Additionally, we are building completely new integrated resources that will bring together various datasets. All of these developments will be seamlessly integrated into the Digital Diversity Hub, providing a unified platform for accessing and utilizing our data.