Empirical Null Estimation using Discrete Mixture Distributions and its Application to Protein Domain Data

TL;DR

This paper introduces data-driven methods for estimating the empirical null distribution in mutation studies using discrete mixture models, enabling more accurate identification of significant mutations in protein domain data.

Contribution

It proposes novel procedures for empirical null estimation with discrete mixtures and applies them to large-scale mutation inference in protein domains.

Findings

Effective cut-off determination methods for null distribution.

Improved mutation significance detection in protein data.

Validation on simulated and real datasets.

Abstract

In recent mutation studies, analyses based on protein domain positions are gaining popularity over gene-centric approaches since the latter have limitations in considering the functional context that the position of the mutation provides. This presents a large-scale simultaneous inference problem, with hundreds of hypothesis tests to consider at the same time. This paper aims to select significant mutation counts while controlling a given level of Type I error via False Discovery Rate (FDR) procedures. One main assumption is that there exists a cut-off value such that smaller counts than this value are generated from the null distribution. We present several data-dependent methods to determine the cut-off value. We also consider a two-stage procedure based on screening process so that the number of mutations exceeding a certain value should be considered as significant mutations. Simulated and protein domain data sets are used to illustrate this procedure in estimation of the empirical null using a mixture of discrete distributions.