A Model for Distributed Processing and Analyses of NGS Data under Map-Reduce Paradigm

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Next-Generation Sequencing (NGS) technologies have led to an exponential increase in sequencing data volume.
• The rapid growth of NGS data presents significant challenges in storage, integration, processing, and analysis.
• Existing models struggle to efficiently manage and analyze the scale of data generated by modern sequencing.

Purpose of the Study:

• To propose a novel distributed model using the Map-Reduce paradigm to address the challenges of NGS big data.
• To develop a modularized, multi-phase architecture for efficient processing and analysis of genomic data.
• To enhance the integration and analysis of individual and multiple genomic profiles.

Main Methods:

• A distributed model based on the Map-Reduce paradigm was designed.
• The model employs a modularized approach with three distinct phases for data processing.
• Phase 1 granulates alignment data for base-level genomic information; Phases 2 and 3 process this in parallel for integrated DNA profiles and individual contigs, respectively.

Main Results:

• The proposed model effectively generates detailed base-level genomic information.
• Parallel processing in later phases creates integrated DNA profiles and individual contigs.
• Simulated and real experimental results demonstrate the model's effectiveness and superiority over existing tools.

Conclusions:

• The novel distributed Map-Reduce model provides an effective solution for managing and analyzing NGS big data.
• The modular architecture facilitates diverse analytical pipelines and generates valuable genomic repositories.
• The model shows significant potential for various applications in genomic data analysis and research.