Infringement of Individual Privacy via Mining Differentially Private GWAS Statistics
Individual privacy in genomic era is becoming a growing concern as more individuals get their genomes sequenced or genotyped. Infringement of genetic privacy can be conducted even without raw genotypes or sequencing data. Studies have reported that summary statistics from Genome Wide Association Studies (GWAS) can be exploited to threat individual privacy. In this study, we show that even with differentially private GWAS statistics, there is still a risk for leaking individual privacy. Specifically, we constructed a Bayesian network through mining public GWAS statistics, and evaluated two attacks, namely trait inference attack and identity inference attack, for infringement of individual privacy not only for GWAS participants but also regular individuals. We used both simulation and real human genetic data from 1000 Genome Project to evaluate our methods. Our results demonstrated that unexpected privacy breaches could occur and attackers can derive identity information and private information by utilizing these algorithms. Hence, more methodological studies should be invested to understand the infringement and protection of genetic privacy.
KeywordsBayesian Network Genome Wide Association Study Inference Attack Differential Privacy Private Statistic
The work is supported in part by US National Science Foundation (DGE-1523115 and IIS-1502273 to XW, and DGE-1523154 and IIS-1502172 to XS).
- 5.The Health Insurance Portability and Accountability Act of 1996 (HIPAA). http://www.hhs.gov/hipaa/
- 6.Shi, X., Wu, X.: Genetic privacy: risks, ethics, and protection techniques. In: The Workshop on Data Science Learning and Applications to Biomedical and Health Sciences, pp. 57–62, New York, NY (2016)Google Scholar
- 7.Homer, N., Szelinger, S., Redman, M., Duggan, D., Tembe, W., Muehling, J., Pearson, J.V., Stephan, D.A., Nelson, S.F., Craig, D.W.: Resolving individuals contributing trace amounts of DNA to highly complex mixtures using high-density SNP genotyping microarrays. PLoS Genet. 4(8), e1000167 (2008)CrossRefGoogle Scholar
- 9.Wang, R., Li, Y.F., Wang, X., Tang, H., Zhou, X.: Learning your identity and disease from research papers: information leaks in genome wide association study. In: 16th ACM Conference on Computer and Communications Security, pp. 534–544. ACM (2009)Google Scholar
- 12.Wang, Y., Wu, X., Shi, X.: Using aggregate human genome data for individual identification. In,: IEEE International Conference on Bioinformatics and Biomedicine, pp. 410–415. IEEE, Shenzhen, China (2013)Google Scholar
- 13.Hindorff, L.A., MacArthur, J., Morales, J., Junkins, H.A., Hall, P.N., Klemm, A.K., Manolio, T.A.: A Catalog of Published Genome-wide Association Studies. http://www.genome.gov/gwastudies
- 14.Fienberg, S.E., Slavkovic, A., Uhler, C.: Privacy preserving GWAS data sharing. In: 11th International Conference on Data Mining Workshops, pp. 628–635. IEEE (2011)Google Scholar
- 15.Johnson, A., Shmatikov, V.: Privacy-preserving data exploration in genome-wide association studies. In: 19th ACM International Conference on Knowledge Discovery and Data Mining, pp. 1079–1087. ACM, Chicago, IL (2013)Google Scholar
- 18.Bhaskar, R., Laxman, S., Smith, A., Thakurta, A.: Discovering frequent patterns in sensitive data. In: 16th ACM International Conference on Knowledge Discovery and Data Mining, pp. 503–512. ACM, Washington, DC (2010)Google Scholar
- 19.Chaudhuri, K., Monteleoni, C.: Privacy-preserving logistic regression. In: 23rd Annual Conference on Neural Information Processing Systems, pp. 289–296. Citeseer, Vancouver, B.C., Canada (2008)Google Scholar
- 20.Kifer, D., Machanavajjhala, A.: No free lunch in data privacy. In: 17th ACM International Conference on Knowledge Discovery and Data Mining, pp. 193–204. ACM, San Diego, CA (2011)Google Scholar
- 21.Lee, J., Clifton, C.: Differential identifiability. In: 18th ACM International Conference on Knowledge Discovery and Data Mining, pp. 1041–1049. ACM, Beijing, China (2012)Google Scholar