Associate Professor Masanori NOJIMA
TEL: +81-3-6409-2340
E-mail: nojima{at}ims.u-tokyo.ac.jp
Lab HP
【Key Words】 biostatistics, medical research design, big data, clinical trials, clinical epidemiology, epigenetics, bioinformatics, DNA methylation, machine learning, social medicine
Research activities in our laboratory: A lot of works are published as collaborative research by being in charge of research design and statistical analysis in clinical trials or clinical epidemiological research. Analyzing confounding and correlations between variables, creating prediction models, we always try to find true relationships by finding and correcting pseudo correlation, and create statistical models useful for medicine and society. We are also involved in environmental epidemiology research and epidemiological research with multi-omics data, and incorporates analyses using machine learning methods such as LASSO and Random Forest in addition to the basic analysis method used in biostatistics. This makes it possible to predict more accurately than conventional methods while maintaining simplicity. We are attempting to predict the onset of future diseases for specific diseases from various environmental pollutants and metabolic products in the blood.
Another important interest of us is epigenetics. In particular, we have studied DNA methylation, using large-scale data. In a recent study, integrating DNA methylation microarray data, gene expression arrays data, and micriRNA expression data, we summarized gene expression control system using statistical models, and reported epigenetic changes and its importance in cancer tissues (Nojima et. al, Mol Cancer 2016). Moreover, we are also involved in social medicine research based on questionnaire survey (e.g. nursing research field). We always try to provide easy-to-understand analyses for collaborators and readers of publications by using simple model, and promote new medical development from the standpoint of “statistics”.
