A semiparametric mixture method for local false discovery rate estimation

TL;DR

This paper introduces a semiparametric mixture model combining Efron's empirical null and log-concave density estimation to improve local false discovery rate estimation, especially in high-dimensional settings.

Contribution

It presents a novel, flexible approach that extends existing methods to high-dimensional data for more accurate false discovery rate estimation.

Findings

Outperforms existing methods in simulations

Effective in high-dimensional scenarios

Demonstrated with astronomy and microarray case studies

Abstract

We propose a semiparametric mixture model to estimate local false discovery rates in multiple testing problems. The two pilars of the proposed approach are Efron's empirical null principle and log-concave density estimation for the alternative distribution. Compared to existing methods, our method can be easily extended to high dimension. Simulation results show that our method outperforms other existing methods and we illustrate its use via case studies in astronomy and microarray.