Using Ethereum blockchain to store and query pharmacogenomics data via smart contracts

Area of Science:

• Bioinformatics
• Blockchain Technology
• Genomic Data Management

Background:

• Pharmacogenomics data is crucial for clinical decisions, necessitating secure storage and sharing protocols.
• Ethereum smart contracts present a promising solution for high-integrity, secure data management.
• The 2019 iDASH competition focused on developing efficient Ethereum smart contracts for gene-drug relationship data.

Purpose of the Study:

• To design and evaluate Ethereum smart contracts for efficient storage and querying of pharmacogenomics data.
• To develop a scalable and time-efficient solution for managing gene-variant-drug interactions on the blockchain.
• To compare the performance of different smart contract designs for pharmacogenomics data management.

Main Methods:

• Developed an index-based, multi-mapping smart contract to store gene-variant-drug triplets with outcomes, searchable by unique identifiers.
• Created an optimized 'fastQuery' smart contract that consolidates identical gene-variant-drug combinations for improved scalability.
• Evaluated both contract solutions on a private, proof-of-authority network, testing insertion and query performance with varying database sizes.

Main Results:

• Both challenge and fastQuery solutions demonstrated near-linear memory and time usage for small databases (<1,000 entries).
• The 'fastQuery' contract maintained linear scaling for larger databases (1000-10,000 entries) and showed significant improvements in scalability and query efficiency.
• The 'fastQuery' solution offered a 2-fold improvement in query time and a 10-fold improvement in memory usage compared to the challenge solution for 2-AND queries.

Conclusions:

• Ethereum blockchain technology can efficiently store and query pharmacogenomics data.
• The developed smart contract solutions offer a viable method for managing sensitive clinical data with high integrity and efficient access.
• These blockchain-based solutions have the potential to be extended for storing diverse clinical data and other fields requiring secure, accessible data management.