Making scientific content computable
The Applied Semantics research group focuses on the following major topics:
- Making data FAIR (Findable, Accessible, Interoperable and Reusable)
- Shared Semantics
- Knowledge Discovery
- Cause-effect Mechanistic Knowledge Graphs
The big data paradigm is highly relevant to the biomedical field with its ever increasing growth of scientific publications – and omics data. However, it is often a challenge to capture and organize relevant scientific knowledge from unstructured text in literature. When it comes to data, problems often arise with insufficient reusability or low quality of the data.
The Applied Semantics group makes both data and knowledge computable and analyzable through data curation, standardization and data management. Our group has expertise in shared semantics, which is the basis for interoperability of data and knowledge. We can achieve semantic interoperability by relating biological entities to standard terminologies or ontologies. Ontologies and terminologies also serve as the basis for semantic based knowledge discovery systems.
Another focus of the group is on knowledge graphs which consist of cause-and-effect mechanisms extracted from scientific publications. Knowledge graphs are multimodal, disease-specific and comprise biological entities ranging from the genetic level to the phenotypic level together with various relationships among them. Our current knowledge graphs cover neurodegenerative, psychiatric and metabolic disorders.
Models that integrate data and knowledge form the basis of our approaches to precision medicine. They allow us to analyze patient-level data taking into account the state of knowledge about disease mechanisms. Some of the main applications resulting from the combination of knowledge graphs and data are the stratification of patients and the identification of comorbidities based on mechanisms, personalized medicine, and drug repurposing.