Secure large-scale genome data storage and query.

Area of Science:

• Bioinformatics
• Cloud Computing
• Database Security

Background:

• Cloud computing offers scalable, cost-effective infrastructure for big data science, including large-scale genome data analysis.
• Privacy and security concerns hinder the adoption of public cloud for sensitive genomic data.
• Developing secure genome databases with encrypted data deposition and query capabilities is essential.

Purpose of the Study:

• To propose a novel, secure mechanism for handling count queries on encrypted genome data within a cloud environment.
• To address the challenge of balancing data security with system speed and scalability for real-world genomic data demands.
• To enhance the efficiency of secure query execution on large genomic datasets.

Main Methods:

• Implementation of a secure count query mechanism on an open-source graph database (Neo4j).
• Development of a new tree indexing method to overcome performance bottlenecks in existing approaches.
• Evaluation using a real-world dataset of approximately 735,317 Single Nucleotide Polymorphisms (SNPs).

Main Results:

• The proposed method significantly reduces query execution time compared to existing techniques.
• An arbitrary count query on a 212 GB dataset is executed in under one minute.
• This represents a substantial improvement over the best-known algorithm, which takes around 7 minutes.

Conclusions:

• The developed framework demonstrates the feasibility of using graph databases for secure storage of large-scale genome data in untrusted environments.
• The underlying crypto-system and security assumptions are well-suited for genomic data use cases and can be generalized.
• This research paves the way for more secure and efficient cloud-based genomic data analysis.