Privacy-preserving data exploration in genome-wide association studies

Abstract

Genome-wide association studies (GWAS) have become a popular method for analyzing sets of DNA sequences in order to discover the genetic basis of disease. Unfortunately, statistics published as the result of GWAS can be used to identify individuals participating in the study. To prevent privacy breaches, even previously published results have been removed from public databases, impeding researchers’ access to the data and hindering collaborative research. Existing techniques for privacy-preserving GWAS focus on answering specific questions, such as correlations between a given pair of SNPs (DNA sequence variations). This does not fit the typical GWAS process, where the analyst may not know in advance which SNPs to consider and which statistical tests to use, how many SNPs are significant for a given dataset, etc. We present a set of practical, privacy-preserving data mining algorithms for GWAS datasets. Our framework supports exploratory data analysis, where the analyst does not know a priori how many and which SNPs to consider. We develop privacy-preserving algorithms for computing the number and location of SNPs that are significantly associated with the disease, the significance of any statistical test between a given SNP and the disease, any measure of correlation between SNPs, and the block structure of correlations. We evaluate our algorithms on real-world datasets and demonstrate that they produce significantly more accurate results than prior techniques while guaranteeing differential privacy.

Differential Privacy with Bounded Priors

“…To assess the potential gain in utility of our relaxation, we focus on a particular application of DP, by re-evaluating the privacy protecting mechanisms in genome-wide association studies [10,19,22] for the release of SNPs with high χ 2 -statistics. Our results show that, for a bounded adversarial model, we require up to 2500 fewer patients in the study, in order to reach an acceptable tradeoff between privacy and medical utility.…”

“…For instance, Fredrikson et al [6] investigate personalized warfarin dosing and demonstrate that for privacy budgets effective against a certain type of inference attacks, satisfying DP exposes patients to highly increased mortality risks. Similarly, studies on privacy in genome-wide association studies (GWAS) [10,19,22] consider differential privacy as a protective measure against an inference attack discovered by Homer et al [9,20]. These works show that for reasonably small values of , the medical utility is essentially null under DP, unless there is an access to impractically large patient datasets.…”

GenoGuard: Protecting Genomic Data against Brute-Force Attacks

“…p m is a non-uniform sequence distribution and we assume γ ≤ 3 − √ 5 ≈ 0.76, which is a requirement for Corollary 1 (in [13]). This assumption is reasonable considering the length of the sequence n (≥ 20000) 5 . To estimate γ, we can consider the sequence with all major alleles and pessimistically assume each major allele frequency is 0.995, large enough to give an upper bound for real datasets.…”

“…As a response to the above privacy breach in published genomic statistics, Fienberg et al [4] propose to apply Laplacian noise to the released data to achieve differential privacy. Another approach to achieving differential privacy in genome-wide association study is proposed by Johnson and Shmatikov [5]. Yu et al [6] present scalable privacy-preserving methods in genome-wide association studies based on Laplace mechanism and exponential mechanism.…”

“…Existing mechanisms for protecting the privacy of genomic data include (i) anonymization, which has proven to be ineffective for genomic data [2], [3], (ii) adding noise to published genomic data or statistics for medical research (e.g., to guarantee differential privacy [4], [5], [6]), (iii) computation partitioning [7], and (iv) cryptography (e.g., homomorphic encryption [8], [9], private set intersection [10], etc.). In this work, we focus mainly on the personal use of genomic data, such as healthcare or DTC services.…”