Comparison of DNA methylation measured by Illumina 450K and EPIC BeadChips in blood of newborns and 14-year-old children.

TitleComparison of DNA methylation measured by Illumina 450K and EPIC BeadChips in blood of newborns and 14-year-old children.
Publication TypeJournal Article
Year of Publication2018
AuthorsSolomon O, MacIsaac J, Quach H, Tindula G, Kobor MS, Huen K, Meaney MJ, Eskenazi B, Barcellos LF, Holland N
JournalEpigenetics
Pagination1-10
Date Published2018 Aug 15
ISSN1559-2308
Abstract

Analysis of DNA methylation helps to understand the effects of environmental exposures as well as the role of epigenetics in human health. Illumina, Inc. recently replaced the HumanMethylation450 BeadChip (450K) with the EPIC BeadChip, which nearly doubles the measured CpG sites to >850,000. Although the new chip uses the same underlying technology, it is important to establish if data between the two platforms are comparable within cohorts and for meta-analyses. DNA methylation was assessed by 450K and EPIC using whole blood from newborn (n = 109) and 14-year-old (n = 86) participants of the Center for the Health Assessment of Mothers and Children of Salinas. The overall per-sample correlations were very high (r >0.99), although many individual CpG sites, especially those with low variance of methylation, had lower correlations (median r = 0.24). There was also a small subset of CpGs with large mean methylation β-value differences between platforms, in both the newborn and 14-year datasets. However, estimates of cell type proportion prediction by 450K and EPIC were highly correlated at both ages. Finally, differentially methylated positions between boys and girls replicated very well by both platforms in newborns and older children. These findings are encouraging for application of combined data from EPIC and 450K platforms for birth cohorts and other population studies. These data in children corroborate recent comparisons of the two BeadChips in adults and in cancer cell lines. However, researchers should be cautious when characterizing individual CpG sites and consider independent methods for validation of significant hits.

DOI10.1080/15592294.2018.1497386
Alternate JournalEpigenetics
PubMed ID30044683

McGill Logo

CIUSSS ouest montreal logo

 



  • Douglas Hospital
  • Dobell Pavillion
  • Brain imaging centre