This work package will integrate drug-target-disease inter-relationships through molecular data using a knowledge graph.
We will use innovative AI algorithms and data from many public repositories on chemical structure, biology and health informatics to predict development, progression of, and outcome from multimorbidity. We will examine the effect of specific drug treatments in amplifying or reducing risk and will look across the different types of evidence to check consistency to ensure our findings are as sound as they can be. Bringing all this data into a knowledge graph enables artificial intelligence agents to be trained on these curated and validated relationships between drugs, target proteins and diseases. This in turn enables further prediction of unknown links and further study of known relationships.