Fig. 9From: LuxRep: a technical replicate-aware method for bisulfite sequencing data analysisAUROCs of differential methylation calls. Accuracy in determining differential methylation was measured by generating datasets consisting of two groups (A and B) with varying \(\Delta \theta\) (\(\theta _A\) and \(\theta _B\) levels are shown in top panels and when one or two of three replicates have low \(\text{BS}_{\rm eff}\) (‘GGB’ and ‘GBB’, respectively.) For ‘GBB’ (top box) and ‘GGB’ (bottom box) \(N_{\rm{BS}} = 10\) and \(N_{\rm{BS}} = 6\), respectively, for each technical replicate. X-axis shows whether HMC or ADVI was used to evaluate or approximate the posteriorsBack to article page