ReadDB provides efficient storage for mapped short reads.

Area of Science:

• Bioinformatics
• Genomics
• Computational Biology

Background:

• High-throughput sequencing generates massive datasets (over 10^8 reads per experiment).
• Existing tools inadequately address storage and retrieval challenges for large aligned sequencing datasets.
• Efficient data management is crucial for analyzing cellular functions through sequencing.

Purpose of the Study:

• To introduce ReadDB, a network-accessible column store database system.
• To provide a solution for storing and retrieving large collections of aligned high-throughput sequencing data.
• To facilitate visualization and analysis of genomic interval data.

Main Methods:

• ReadDB is implemented as a network server.
• It stores aligned read positions and responds to queries on genomic intervals.
• Provides either contained reads or histogram-based interval summaries.

Main Results:

• ReadDB demonstrates high performance on datasets ranging from 10^5 to 10^8 reads.
• Performance is within a factor of two of local-storage methods.
• Outperforms other network-based methods by three to five times.

Conclusions:

• ReadDB serves as a high-performance foundation for ChIP-Seq and RNA-Seq analysis.
• The client-server model enables convenient access without shared network file systems or large local storage.
• Offers a new method for storing genome-aligned reads, optimized for applications not requiring read sequence or mismatch data.