Fig. 2From: Simulating ComBat: how batch correction can lead to the systematic introduction of false positive results in DNA methylation microarray studies“mod” refers to a model matrix for the outcome of interest. a Boxplots of p-value distributions without added batch effects and under various conditions. Dotted line indicates expected mean p-value of 0.5. b Boxplots of p-value distributions with simulated batch effects and under various conditions. c Boxplots of FDR false significant sites without batch. d Boxplots of FDR false significant sites without batch. Boxplots for “without ComBat” are indicating the false positive sites due to uncorrected batch effects for the respective sample distributionsBack to article page