Responsible Genomic Data Sharing

Challenges and Approaches

Xiaoqian Jiang, Haixu Tang

• Publisher: Academic Press
• Year: 2020
• Print ISBN: 9780128161975
• eBook ISBN: 9780128163399
• Topic: Biological Sciences
• Subtopic: Genetics & Genomics
• Index: Biological Sciences

Responsible Genomic Data Sharing: Challenges and Approaches brings together international experts in genomics research, bioinformatics and digital security who analyze common challenges in genomic data sharing, privacy preserving technologies, and best practices for large-scale genomic data sharing. Practical case studies, including the Global Alliance for Genomics and Health, the Beacon Network, and the Matchmaker Exchange, are discussed in-depth, illuminating pathways forward for new genomic data sharing efforts across research and clinical practice, industry and academia.- Addresses privacy preserving technologies and how they can be applied to enable responsible genomic data sharing- Employs illustrative case studies and analyzes emerging genomic data sharing efforts, common challenges and lessons learned- Features chapter contributions from international experts in responsible approaches to genomic data sharing

Section II

Privacy-preserving techniques for responsible genomic data sharing

Outline

Chapter 5. Homomorphic encryption

Chapter 6. Secure multi-party computation

Chapter 7. Game theory for privacy-preserving sharing of genomic data

Chapter 8. Trusted execution environment with Intel SGX

Chapter 5

Homomorphic encryption

Kim Laine, PhD Microsoft Research, Redmond WA, United States

Abstract

Homomorphic encryption is a powerful encryption technology that allows computations to be performed directly on encrypted data, without requiring access to a secret key. Since its inception in 2009, the technology has seen a huge number of changes, improving its performance by multiple orders of magnitude compared to the first implementations. Today's cloud computing economies make homomorphic encryption more relevant than ever, as it can be used to build completely private outsourced storage and computation services. In this chapter we discuss many fundamental aspects of homomorphic encryption: schemes and constructions, security, implementations, performance, and applications, specifically emphasizing applications in the biomedical field.

Keywords

Cryptography; Homomorphic encryption

1. Overview

A fundamental challenge in storing and processing of biomedical data is the high level of data privacy that handling of such data necessitates. These requirements are sometimes enforced by governments and sometimes stem from liability concerns or fear of negative publicity from data breaches. While public cloud services could in principle provide the needed storage and computation infrastructure, private patient data should not be exposed to the cloud service operators. A simple partial remedy is for the hospital to store only encrypted data in the cloud, but this significantly weakens the capabilities of the service: any computation or operation on the data requires it to be downloaded back to the hospital's system, to be decrypted, and then computed on. Another common solution is to simply give the cloud service access to the decryption key and trust that they implement appropriate access policies to prevent anyone from decrypting all of the data and downloading or sharing it.

This is one of the reasons why hospitals and medical clinics prefer to store their data on premises, rather than outsourcing storage and computation to much more cost-efficient and flexible public cloud services. However, it is extremely costly and challenging even for large hospitals with well-funded IT departments to operate their own data centers securely and re...

Table of contents

• Cover image
• Title page
• Table of Contents
• Copyright
• Contributors
• Section I. Privacy challenges in genomic data sharing
• Section II. Privacy-preserving techniques for responsible genomic data sharing
• Index