FDR functions for permutation-based estimators, including pi0 as well as FDR confidence intervals. The confidence intervals account for dependencies between tests by the incorporation of an overdispersion parameter, which is estimated from the permuted data. Also included are options for an analog parametric approach.
