Whole blood was collected from each previously recruited patient during clinic visits in vials containing EDTA. Peripheral blood mononuclear fractions were isolated using density centrifugation with Ficoll/Paque (GE Healthcare Life Sciences, Piscataway, NJ, USA). CD4+ T cell were isolated using MACS magnetic bead CD4+ T-cell isolation kit (Miltenyi Biosystems, Auburn, CA). Genomic DNA was isolated from the CD4+ T cells using Qiagen DNEasy kit (Qiagen, Germantown, MD, USA). 250ng of DNA from each sample was bisulfite converted using the EZ-96 DNA Methylation Kit (Zymo, Irvine, CA, USA) following the manufacturer’s instructions.
Cy5 and Cy3
Patient 22 (Female)
DNA methylation data analysis was performed in the R statistical computing environment (v.3.6.3). Raw .idat files were generated for each sample and underwent quality control and downstream analysis in the R package minfi (v.1.32.0). The following probes were removed based on best practices as used in DNAmArraypackage (v.0.1.1): probes with less than three beads and zero intensity values across all samples. Thereafter, correction for background signal and dye bias and normalization of signal intensities using functional normalization in the preprocessFunnorm.DNAmArray function was done. The first three principal component values, calculated from signal intensities of control probes present on all array spots to correct for technical variation, were utilized. More probes were removed, which included the following: probes with detection P-values < 0.01 and probes that returned signal intensities in fewer than 98% of samples. Remaining probes’ signal intensities were converted to M-values with limits of ±16. To reduce the number of probes with potential technical issues, we followed the criteria described by Zhou, Laird & Shen (2017) to mask any probes: a unique probe sequence of less than 30bp, mapping to multiple sites in the genome, polymorphisms that cause a color channel switching in type I probes, inconsistencies in specified reporter color channel and extension base, mapping to the Y chromosome, and/or having a polymorphism within 5bp of the 3’ end of the probe with a minor allele frequency (MAF) > 1% with exception of CpG-SNPs with C>T polymorphisms which we retained for analysis. ComBatfunction in the sva (v.3.34.0) package was used to perform batch correction. Average Beta