Scalable De Novo Genome Assembly Using a Pregel-Like Graph-Parallel System.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• De novo genome assembly reconstructs genomes from short DNA sequences without a reference.
• High-throughput sequencing necessitates efficient assembly tools for rapid genome discovery.

Purpose of the Study:

• To introduce PPA-assembler, a toolkit for distributed de novo genome assembly.
• To provide a scalable and efficient solution for assembling large genomic datasets.

Main Methods:

• Utilizes the de Bruijn graph approach for sequence assembly.
• Implements operations as programs within Google's Pregel framework for distributed computing.
• Designed for deployment on generic clusters.

Main Results:

• Demonstrates significantly higher efficiency compared to state-of-the-art methods on large real and simulated datasets.
• Achieves comparable genome sequencing quality to existing approaches.
• PPA-assembler offers strong performance guarantees for its operations.

Conclusions:

• PPA-assembler provides an efficient and scalable solution for de novo genome assembly in a distributed environment.
• The toolkit accelerates the discovery of new genomes through improved assembly performance.
• PPA-assembler is open-sourced to facilitate broader research adoption.