Cross-Cohort Modeling of Microbiome–Phenotype Associations for Privacy Risk Assessment

We aim to assess the extent to which microbiome data can be used for privacy risk identification to infer sensitive genotype or phenotype information. We will identify associations between microbes and phenotypes to quantify the privacy risk of microbiome data sharing. We will investigate different human microbiome components and identify microbiome features that may carry higher inference sensitivity in order to inform privacy-preserving data sharing strategies. Through this work, we hope to contribute to a future where microbiome data sharing and analysis will be more secure and ethical.