The outstanding quality and extent of population biobanks and health registries in Norway ensures a solid foundation for high quality research.
In this project we aim to combine information about health and disease from three of the large population based health studies, with health registries to better understand underlying biological mechanisms and risk factors for disease.
Of particular interest is the relationship between genes and the environment. To investigate this we will study health in relation to genetic information from approximately 80,000 individuals whom have participated in either the Nord-Trøndelag Health Study or The Tromsø Study, in addition to associations between genes, environment and cancer in The Norwegian Women and Cancer Study.
We need improved statistical methods and models to exploit the opportunities within these large-scale data sets. An important part f our project is therefore to develop such methods. We have focused on three specific topics; we have developed a test for high-dimensional mediation, we have studied different methods for integration of several sources of high-dimensional data and developed an ensamble-based method for extracting information from these sources, and we have developed a robust method for variable screening in so-called ultra-high dimensional regression problems. Related to this we are also working on a robustified version of one method for dataintegration, to handle situations with outliers, and finally we have studied a new method for dealing with multiple testing problems related to the mediation problem above.
Another important part of our project is to establish a structured training and interdisciplinary forum for researchers in biostatistics and bioinformatics. This training is aimed at Postdoctoral fellows who wish to establish their own independent research careers. We are well underway with competence exchange and discussion groups.
As an increasing amount of health data becomes available for research, it is important for the research activities to be cognizant of public opinion. Ethical training is therefore an important part of our effort.
We present an ambitious but realistic three year proposal to leverage existing digital biologic information from three of the largest prospective cohort studies in Norway, enriched with linkages to a comprehensive list of health registries, to better understand the biology for health and disease within diverse disease domains.
We will build national methodological competence and capacity in the analysis of large-scale biobanks and health registries by focusing on relevant methodological developments.
We will investigate complex ethical questions on the horizon, such as gene-based follow up of participants, in order to secure that the interest of large-scale biobank research and the participants are harmonious and compatible with an ethical commitment to the principle of reciprocity.
We will expand our strong interdisciplinary research team within our three partner universities in Norway, and continue to collaborate with world leaders in biostatistics and bioinformatics.
We will use this foundation to mentor and train six postdoctoral fellows at the multidisciplinary interface of applied and methodologically driven biostatistics and bioinformatics to meet the increasing complexity of the new era of precision medicine.
Our proposal is in compliance with the overarching strategies of the partner universities to build strong statistical and bioinformatics expertise and capacity, and the proposal will be in synergy with our ongoing RCN and NIH supported efforts.