Replication and meta-analyses
Genome-wide significant CpG sites that passed Bonferroni correction (P< 1.14×10-7, which is 0.05 / 436,824) in whole blood DNA were selected for replication in two independent cohorts, Lifelines and EGEA. Lifelines did not include an assessment of AHR, so only included a ClinR phenotype available for replication. The weighted Z-score method was used to meta-analyze the results of discovery and replication cohorts, considering EGEA used methylC-capture sequencing method, which was different from discovery and Lifelines (450K array), to investigate DNA methylation. CpG sites that passed the epigenome-wide significance threshold of P< 1.14×10-7 (Bonferroni correction) in the meta-analysis of all studies were considered to be replicated. Finally, top sites identified in whole blood were verified in nasal brushes.