Preserving genome privacy in research studies

As the cost of genome sequencing continues to fall, whole genome sequencing data have become a viable alternative for improving diagnostic accuracy and supporting personalized medicine. Although they have the potential to advance public health and accelerate scientific discoveries, massive collections of genomic data also raise significant concerns about individual privacy. Like traditional clinical information, human genomes may reveal information about individuals (e.g., identity, ethnic group, disease association, predisposition to diseases such as diabetes or cancer, etc.) Even more concerning is the fact that the information is shared with ancestors and descendants, and thus loss of privacy may put the privacy of the entire family at risk. Genome privacy is a big challenge for the entire biomedical community, particularly since scientific discoveries depend on data sharing and obfuscation of data is not a good option to protect privacy. Multiple factors are involved in genomic privacy research. The components that can be used to better protect genome privacy include, but are not limited to, legal, ethical and technical aspects, e.g., federal laws, policies and regulations, informed consent policies, data use agreements, secure data repositories, as well as privacy-preserving data analysis methods. However, genome privacy challenges cannot be addressed by any single component alone. We envision that better privacy protection can be achieved through the incorporation of multiple components. The goal of this chapter to introduce the state-of-the-art in genome privacy research. This chapter begins with an introduction of genome privacy followed by an overviewof the legal, ethical and technical aspects of genome privacy. After formalizing the genome privacyproblem, we will review existing attack models on genomic data. The techniques for mitigating these attacks are discussed. This chapter concludes with the discussion of the challenges and the future directions in genome privacy research.