How to make DNA methylome wide association studies more powerful

Abstract

Genome-wide association studies had a troublesome adolescence, while researchers increased statistical power, in part by increasing subject numbers. Interrogating the interaction of genetic and environmental influences raised new challenges of statistical power, which were not easily bested by the addition of subjects. Screening the DNA methylome offers an attractive alternative as methylation can be thought of as a proxy for the combined influences of genetics and environment. There are statistical challenges unique to DNA methylome data and also multiple features, which can be exploited to increase power. We anticipate the development of DNA methylome association study designs and new analytical methods, together with integration of data from other molecular species and other studies, which will boost statistical power and tackle causality. In this way, the molecular trajectories that underlie disease development will be uncovered.

Keywords: DNA methylation; DNA methylation association studies (methWAS); developmental trajectories; epigenetic epidemiology; epigenome-wide association study (EWAS); gene × environment (GxE).

Figure 1.. Simplified DNA methylation trajectories for a subject without (green solid line) and with disease (dotted blue line) where (A & D) methylation changes as a consequence of disease, or (B & E) disease occurs as a consequence of methylation, or (C & F) there is no causal relationship between disease and methylation (a confounding factor independently affects both disease status and methylation), respectively.

T1 and Td represent times when DNA sample is collected and disease occurred, respectively (these events are also represented with a square and ‘D’). Tc is the time when a confounding factor (e.g., environmental exposures) affects methylation (denoted with a ‘C’). In the top panel (A–C), DNA sample is collected after disease onset, we observe a positive association between methylation and disease, but cannot distinguish between each of the three scenarios D->M, M-> D and confounding. In the bottom panel (D–F), DNA sample was obtained prior to disease onset, allowing us to rule out D -> M, but both M -> D and confounding are still possibilities.