Small-scale inference: Empirical Bayes and confidence methods for as few as a single comparison

TL;DR

This paper demonstrates that local false discovery rate (LFDR) estimation can be effectively performed with minimal data, such as a single comparison, by constraining the null hypothesis proportion, improving inference accuracy in proteomics.

Contribution

It introduces a method for estimating LFDR with as few as one comparison by restricting the null hypothesis proportion, enhancing inference in small-scale studies.

Findings

Estimated null hypothesis proportions: ~20% in group I, ~90% in group II.

LFDR estimates align well with confidence levels in group I, and closely with LFDR in group II.

Simulation results show confidence and LFDR methods perform better at low and high null proportions.

Abstract

By restricting the possible values of the proportion of null hypotheses that are true, the local false discovery rate (LFDR) can be estimated using as few as one comparison. The proportion of proteins with equivalent abundance was estimated to be about 20% for patient group I and about 90% for group II. The simultaneously-estimated LFDRs give approximately the same inferences as individual-protein confidence levels for group I but are much closer to individual-protein LFDR estimates for group II. Simulations confirm that confidence-based inference or LFDR-based inference performs markedly better for low or high proportions of true null hypotheses, respectively.