Heterogeneity impacts biomarker discovery for precision medicine

This is the authors' abstract. We don't add key points for this paper.

Precision medicine is advancing patient care for complex human diseases. Discovery of biomarkers to diagnose specific subtypes within a heterogeneous diseased population is a key step towards realizing the benefits of precision medicine. However, popular statistical methods for evaluating candidate biomarkers – fold change (FC) and area under the receiver operating characteristic curve (AUC) – were designed for homogeneous data. Herein, we evaluate the performance of these metrics in heterogeneous populations. Using simulated biomarkers that are nearly ‘ideal’ for distinguishing subgroups of various proportions of the diseased population, we observe that AUC misses all up to subset size of 50% and FC misses all biomarkers entirely. We introduce a simple new measure to address this shortfall and run a series of trials comprised of simulated and biological data to demonstrate its utility for evaluating biomarkers associated with disease subtypes.