Assessing replicability across multiple studies, each examining one or more features; designing optimal multiple testing procedures; inferring causality in observational studies, while accounting for observed and hidden bias; nonparametric testing of independence between random vectors.