DAGGER: A sequential algorithm for FDR control on DAGs
This provides a sequential testing method for structured hypotheses in fields like genomics, though it is incremental as it builds on existing algorithms for special cases.
The authors tackled the problem of controlling false discovery rate (FDR) in multiple testing on directed acyclic graphs (DAGs) with logical constraints, proposing the DAGGER algorithm that guarantees bounded FDR under various dependence structures and shows favorable performance in simulations and a gene ontology dataset.
We propose a linear-time, single-pass, top-down algorithm for multiple testing on directed acyclic graphs (DAGs), where nodes represent hypotheses and edges specify a partial ordering in which hypotheses must be tested. The procedure is guaranteed to reject a sub-DAG with bounded false discovery rate (FDR) while satisfying the logical constraint that a rejected node's parents must also be rejected. It is designed for sequential testing settings, when the DAG structure is known a priori, but the $p$-values are obtained selectively (such as in a sequence of experiments), but the algorithm is also applicable in non-sequential settings when all $p$-values can be calculated in advance (such as variable/model selection). Our DAGGER algorithm, shorthand for Greedily Evolving Rejections on DAGs, provably controls the false discovery rate under independence, positive dependence or arbitrary dependence of the $p$-values. The DAGGER procedure specializes to known algorithms in the special cases of trees and line graphs, and simplifies to the classical Benjamini-Hochberg procedure when the DAG has no edges. We explore the empirical performance of DAGGER using simulations, as well as a real dataset corresponding to a gene ontology, showing favorable performance in terms of time and power.