Fig. 1From: Significant variation in the performance of DNA methylation predictors across data preprocessing and normalization strategiesSchematic overview of the analysis plan to evaluate DNAm algorithm performance. DNAm analyses are conducted using DNAm EPIC array samples in JHS. JHS includes a significant number of technical replicate pairs thereby allowing for a careful investigation of how the removal of unwanted technical variation impacts DNAm algorithm performance across 101 data processing pipelines. JHS has also collected information on disease-related phenotypes, including mortality status after follow-up. This allowed us to assess how the removal of technical variation in DNAm predictor estimates by a data processing pipeline impacts downstream phenotypic association analysesBack to article page