DNA Methylation and Cardiovascular Diseases: A Critical Study on Reliability and Association

Strategic Insights -

Researchers have made a significant contribution to the understanding of DNA methylation and its impact on cardiovascular diseases. A study conducted as part of the Atherosclerosis Risk in Communities (ARIC) study has provided insight into the reliability of DNA methylation measurement using the Illumina Infinium HumanMethylation450 (HM450) BeadChip. The study analyzed DNA methylation patterns in peripheral blood samples from approximately 3,000 African American participants, exploring the association between methylation levels and smoking status.

Key Takeaways:

  • The study used the Illumina Infinium HumanMethylation450 (HM450) BeadChip to measure DNA methylation in peripheral blood samples from approximately 3,000 African American participants.
  • Over 480,000 cytosine-guanine (CpG) dinucleotide sites were surveyed, with 265 technical replicates from 130 participants included to evaluate the impact of technical variation.
  • The researchers calculated the intraclass correlation coefficient (ICC) for each CpG site, ranging between 0 and 1, to compare variation of methylation levels within-and between-replicate pairs.
  • The CpG sites were clustered into low-and high-reliability groups based on posterior probabilities, with ~96% of CpG sites showing genome-wide significant association with smoking status coming from the high reliability cluster.
  • The study suggests excluding CpG sites with low ICC from subsequent association analyses or taking extra caution when associations at such sites are observed.
  • The researchers demonstrated the performance of this clustering when applied to a study of association between methylation levels and smoking status.

Statistics:

  • Approximately 3,000 African American participants were included in the study.
  • Over 480,000 cytosine-guanine (CpG) dinucleotide sites were surveyed on the HM450 BeadChip.
  • 265 technical replicates from 130 participants were included to evaluate the impact of technical variation.
  • The intraclass correlation coefficient (ICC) for each CpG site ranged between 0 and 1.
  • ~96% of CpG sites showing genome-wide significant association with smoking status came from the high reliability cluster.

Sources:

  • Evaluation of microarray-based DNA methylation measurement using technical replicates: the Atherosclerosis Risk in Communities (ARIC) Study. Bmc Bioinformatics, 2014;15():312. (BioMed Central - www.biomedcentral.com/; Bmc Bioinformatics - www.biomedcentral.com/bmcbioinformatics/)
  • M. Bose, Division of Biostatistics, School of Public Health, University of Minnesota, Minneapolis, MN 55455, United States.