Genome-wide analysis of high-dimensional multi-omic data using machine learning approaches specifically designed for “large-p small-n" data, including methods for stable feature selection and addressing multi-collinearity.
Data collection, access and reproducibility
Exploring optimal ways of working with extremely large linked datasets, such as electronic health records, available within secure, privacy-protecting Trusted Research Environments (TREs; also known as Secure Data Environments (SDEs)).
Linking digital footprints data such as shopping data and social media into cohort studies (developing ground truth in smart data research)