Information technologies for clinical pharmacogenetics

A sizable fraction of patients undergoing drug treatment experiences adverse events or a lack of positive effects. A part of this variability in drug response can be explained by genetic differences between patients, which can strongly influence how medications are metabolized and able to bind to their targets the focus of a discipline called pharmacogenetics. However, it is still unclear how pharmacogenetic data can best be integrated, organized and used in routine medical care. The goal of this project is to investigate how web technologies, ontologies and automated reasoning systems can be employed to solve this problem. The project is organized around three driving research questions: 1) How can pharmacogenetic knowledge about common medications be represented in a standardized, ontology- based format that enables automated processing, consistency checking and reasoning? What insights can be gained when such a representation is created? 2) How can individual pharmacogenetic patient data be made available at the point-of-care where electronic health records are unavailable or cannot contain genetic data? Can 2D barcode technologies be used for providing pharmacogenetic data to medical professionals? 3) How can computational clinical decision support be structured and made available through mobile web devices to help medical doctors and pharmacists interpret pharmacogenetic data at the point-of-care? We will conduct an in-depth analysis of existing research and information technologies for pharmacogenetics. Informed by this analysis and promising preliminary results from our own research, we will develop novel formalisms and prototypes based on ontologies and web technologies. Then, we will evaluate these formalisms and prototypes to gain new insights into the IT aspects of pharmacogenetic data. Finally, we will aim to establish a shared formalism for capturing and integrating pharmacogenetic data and rules. The project will be conducted in cooperation with international partners and the Health Care and Life Science Interest Group of the World Wide Web Consortium (W3C).

A sizable fraction of patients undergoing drug treatment experiences adverse drug events or a lack of positive effects. A part of this variability in drug response can be explained by genetic differences between patients, which can strongly influence how medications are metabolized. Being able to use genetic data in clinical practice - an approach commonly referred to as "pharmacogenomics" - could greatly improve the safety and efficacy of many common pharmaceuticals. In this project funded by the FWF, we investigated how novel information technologies could be harnessed to make pharmacogenomics available to all patients around the world.To address this multi-faceted problem space, the project covered a wide range of research topics: How can we represent medical knowledge in computer systems? How should pharmacogenomics testing be deployed and who would profit most from it? How good are currently available genetic tests at capturing the genetic diversity of patients? How can we ensure that patient data are easily available when they are needed during medical care? And finally, how can we use computers to assist medical professionals in using genetic data in daily medical routine?In the project, we devised a system that uses formalized computer languages, so-called ontologies, to represent pharmacogenomic knowledge, demonstrating how this approach can be used to automatically find errors and contradictions in the underlying medical knowledge, and to draw new inferences. We analysed big data sets (e.g., of 1,5 million anonymous patients from Austria and 73 million anonymous patients from the United States) and found that a large fraction of patients and especially elderly patients could profit from pharmacogenomics testing and computer-based decision support. We also documented that many current pharmacogenomic tests and nomenclature systems might in some cases be inadequate for capturing the genetic diversity found in patient populations.To help realize the promises of pharmacogenomics in daily medical practice, we created and tested the Medication Safety Code (MSC) system, which makes medical data and clinical decision support available through mobile systems. Patients can be given specially designed MSC cards that include personal genetic data as quick-response (QR) code. These QR codes can be scanned to yield personalized decision support in a wide variety of health care settings. User evaluations of the MSC system yielded positive results.The work in this project is further translated into practice through a follow-up research project funded by the European Union, where some of the developed methodologies will be studied in practice as part of multinational clinical trial.