Decentralized genomics audit logging via permissioned blockchain ledgering.

Area of Science:

• Bioinformatics
• Computer Science
• Data Security

Background:

• The iDASH Secure Genome Analysis Competition 2018 focused on blockchain-based immutable logging for genomic data access audit trails.
• A key challenge was designing efficient storage and retrieval mechanisms for genomic data access logs using MultiChain 1.0.4.

Purpose of the Study:

• To develop and evaluate blockchain-based immutable logging and querying techniques for cross-site genomic dataset access audit trails.
• To design time/space efficient structures for storing and retrieving genomic data access logs.

Main Methods:

• Utilized the MultiChain stream API, treating it as a key-value store.
• Implemented a two-level index supporting efficient single-clause queries.
• Employed heuristic and binary search for conjunction and timestamp range queries, with complexity independent of dataset size.

Main Results:

• Developed and tested Python implementations of insertion and querying techniques using the Savoir library.
• Validated performance across varying dataset sizes and tested a port to MultiChain 2.0 beta.
• Demonstrated feasibility for genomic query log data with hundreds of megabytes of data and acceptable query times.

Conclusions:

• Presented efficient techniques for storing and querying log data in MultiChain 1.0.4 and 2.0 beta.
• Showcased the viability of permissioned blockchain ledgers for genomic data audit trails.
• Observed a 30%-40% increase in insertion efficiency with MultiChain 2.0 beta compared to version 1.0.4.