Challenges and opportunities for statistical power and biomarker identification arising from rhythmic variation in proteomics

Abstract

Time-of-day variation in the molecular profile of biofluids and tissues is a well-described phenomenon, but—especially for proteomics—is rarely considered in terms of the challenges this presents to reproducible biomarker identification. We provide a case study analysis of human circadian and ultradian rhythmicity in proteins, including in the complement and coagulation cascades and apolipoproteins, with PLG, CFAH, ZA2G and ITIH2 demonstrated as rhythmic for the first time. We also show that rhythmicity increases the risk of Type II errors due to the reduction in statistical power from increased variance, and that controlling for rhythmic time-of-day variation improves statistical power and reduces the chances of Type II errors. We recommend that best practice in proteomics study design should account for temporal variation and that time of sampling be reported as part of study metadata. These simple steps can mitigate against both false and missed discoveries, as well as improving reproducibility.