In clinical research and development of new medicines for diseases such as cancer, Alzheimer’s disease and diabetes, academic research groups and large pharmaceutical companies generate a rapidly increasing amount of data. The data often remain as islands, because it is not possible to link older studies and external databases with the latest experiments. sample-image
HUMIT: From data islands to integrated, networked biomedical data
The Project HUMIT, "Human-centered support of incrementally-interactive data integration using the example of high-throughput processes in the Life Sciences" wants to gain more information from the combination of new and already existing data with new Big Data methods. The project under the leadership of the Fraunhofer Institute for Applied Information Technology FIT in Sankt Augustin addresses both the users in small and large pharmaceutical companies as well as in public research institutions.

The Fraunhofer Institute for Molecular Biology and Applied Ecology (IME), as a pharmaceutical service provider, performs high- throughput experiments for the search of new drugs. In the data already collected, there is much information that could be used for new issues. The German Centre for Neurodegenerative Diseases (DZNE) also collects a considerable amount of experimental data in the research of the causes of Alzheimer’s and Parkinson’s disease, but the material cannot be fully reconciled with the global research databases. Together with Fraunhofer FIT and the laboratory information system provider soventec, the partners want to find ways in which diverse information can be better integrated and used in the future to better understand complex diseases and to more effectively detect the effects and side effects of substances.

The key challenge is the high heterogeneity of biomedical data. Experimental approaches and models are changing in the research very rapidly along with the data structures. The project aims to develop new methods that allow a user to interactively recognize the structure from the existing data and to bring it together with other structures by means of a new technology. Since the integration always only occurs for the necessary data ("as-you-go"), it is easier to check and adjust. The researcher retains the ultimate control over the form and interpretation of its data.

In this way, the project aims to promote data-driven knowledge acquisition. Although the project is mainly dedicated to preclinical research, its attention is also directed to the integration of clinical data and ensures that the high demands on security and privacy of such data are not undermined by Big Data approaches.

There is already strong interest from the pharmaceutical industry and the associated software industry in the future results of the project. The project contributes to maintaining Germany as an attractive research location and to use resources more effectively in health-oriented research.