Methylation array data can simultaneously identify individuals and convey protected health information: an...

Area of Science:

• Genomics
• Bioinformatics
• Medical Informatics

Background:

• Genome-wide methylation arrays are vital for studying complex diseases.
• Current policy allows open access to methylation data, assuming it cannot be linked to personal information.
• This assumption is being challenged by new data analysis techniques.

Purpose of the Study:

• To investigate the potential for extracting personally identifiable information from methylation array data.
• To assess the risk of re-identification for individuals whose data is publicly available.
• To propose policy recommendations for managing sensitive genomic data.

Main Methods:

• Analysis of methylation data from the Illumina HumanMethylation450 array.
• Utilizing 1,069 informative loci to derive genetic information.
• Correlating methylation data with known substance use (alcohol, smoking).

Main Results:

• Genotype information at 1,069 loci was successfully derived from methylation data.
• Alcohol and smoking consumption histories were also extracted from the same data.
• Demonstrated that both identifying and sensitive behavioral data can be obtained simultaneously.

Conclusions:

• Methylation array data contains both potentially identifying information and substance-use histories.
• The risk of re-identification is currently limited due to the reliance on genetic information.
• Recommend restricting access to genome-wide methylation data to approved researchers with data use agreements.