Recent advances in omics technologies are producing data that provide views of biological processes at different resolutions and conditions opening a new era in molecular biology and molecular medicine. An especially exciting aspect of such big data analysis is that it enables novel approaches, methods, and tools for personalized medicine. In particular it is possible to use the data to develop innovative clinical tools that predict either a diagnosis or the best treatment for a patient or a patient group (stratum). However, realizing this potential requires developing novel statistical methods and implementing these for efficient processing of big datasets. In this project we will establish collaboration between research groups with expertise in machine learning and computer system development in order to develop prediction methods and tools for a unique large scale epidemiological study.
As the data collected by the Norwegian Woman and Cancer biobank grow in quantity and variety, needs in terms of machine learning techniques and computational efficiency appeared, leading to new interdisciplinary projects. Interactions therefore begun in the fall of 2014 with Lars Ailo Bongo, computer scientist in Tromsø, as well as with Etienne Birmelé and Vittorio Perduca, statisticians in Paris. Moreover, a PhD funding was obtained for Einar Holsbø on this interdisciplinary topic, also in 2014 (funded by the University of Tromsø). He is co-supervised by Lars Ailo Bongo, Etienne Birmelé and Eiliv Lund.
The proposed collaboration is to improve and develop new interdisciplinary projects that includes research challenges in both statistics and computer science. All four researchers (Birmelé, Bongo, Holsbø and Perduca) are young researchers. We have very limited funds for travel in the new collaboration. This is also the first time after the startup of the collaboration that we can apply for funding from Aurora.