Skip to main content
Fig. 1 | BMC Bioinformatics

Fig. 1

From: Stochastic epigenetic outliers can define field defects in cancer

Fig. 1

Overall strategy for comparing DV algorithms in their ability to identify field defects in cancer: a In a discovery step, we apply five DV algorithms (BT, iEVORA, GAMLSS, J-DMDV, DiffVar) to identify differentially variable CpGs (DVCs) between two normal phenotypes, in two tissue types, as shown. The DV algorithms differ in their sensitivity and control of type-1 error rate, with some DV algorithms not identifying stochastic outlier profiles (i.e. DNA methylation profiles with a few outliers), whilst others are sensitive to them. b In the validation step, we assess the identified DVCs (if any) of each DV algorithm in terms of whether they exhibit progression/enrichment within established neoplastic cells or invasive cancers. This allows an objective comparison of the DV algorithms and helps assess whether stochastic epigenetic outliers identified in step-A using algorithms such as iEVORA are biological or not

Back to article page