Specialization
Biostatistics
Focus of research
My research concerns developing statistical methods for new research designs and new data formats in biomedical science. In particular I focus on omics and big data in general. For integrating omics data I am working on high dimensional multivariate models such as canonical correlation and redundancy analysis and partial least squares methods in general. I developed several software tools that are capable of integrating multiple datasets, each consisting of hundredthousands of variables. Another line of research is focused on using existing registry data for epidemiological cohort studies that are performed in the AMC. I developed methods to jointly perform multiple record linkage and association analyses that are truly unbiased even if the record linkage is done with relatively low quality link variables. This is truly big data analysis because my tools are capable of linking/analyzing datasets covering the entire Dutch population. Both the omics and the linked record data analysis tools are based on parallel computing and make use of clustercomputers, GPU computing and the Dutch grid of computing facilities. Related research subject are related to developing dynamic (prediction) models and causal effects analysis. In addition to these methodological subjects I am co-initiator and co-PI of the HELIUS cohort study of about 25,000 inhabitants of Amsterdam.