DNA was extracted from white blood cells of SKIPOGH participants using a bead-based KingFisher Duo robot extraction system (ThermoFisher, Waltham, Massachusetts), with 1.2 µg of DNA then treated with bisulfite using the EZ DNA Methylation© Kit (Zymo Research). Alternative incubation conditions (described in the EZ DNA Methylation™ Kit bisulfite conversion protocol, point 6 of the appendix) were used when performing the polymerase chain reaction (PCR) using the Illumina Infinium© Methylation Assay, and the final elution was carried out using 8 µl of M-Elution Buffer. DNA methylation levels were then assessed using genome-wide DNA methylation micro-array platforms Illumina HumanBeadChip 450 K (HM450K) (N = 250) and EPIC 850 K (EPIC) (N = 480). Missing CpG data were imputed using the nearest averaging multiple imputation method, beta coefficients were then logit transformed [4 , 5 (link)]. The CPACOR pipeline was used to normalize data and identify quality control issues [6 (link)]. Samples with issues identified in a quality control step, e.g., a call rate of less than 95% were excluded from all analyses (p value < 10–16) (HM450k: N = 9; EPIC: N = 15). CpGs identified as being present across both the EPIC and HM450k arrays (N = 452,453) were subsequently used for all analyses and calculations of epigenetic signatures.
Free full text: Click here