The Benjamini-Hochberg (BH) procedure is a staple of modern high-dimensional data analysis. This method can be made more powerful by incorporating estimators of the number (or proportion) of null hypotheses, yielding an adaptive BH procedure which still controls the false discovery rate (FDR). In this talk we present a unified class of estimators, which encompasses existing and new estimators and which can also be extended to discrete tests. While our focus is on presenting the generality and flexibility of the new class of estimators, we also include some analyses on simulated and real data.
