Nature Methods' note: The articles by Zheng et al. and Rahmani et al. address the suitability of ReFACTor, a tool previously published in Nature Methods by Rahmani et al. (doi:10.1038/nmeth.3809), to correct for cell-type heterogeneity in epigenome-wide association studies. The authors fundamentally differ in their conclusions, but they make their data and analysis software available in the spirit of dialogue and the open discussion of conflicting results. We, and the reviewers of these correspondences, hope that such discussion helps the field move forward. The exchange highlights the need for ground-truth data sets to benchmark current and future tools for epigenomic analysis.