MOCAT

MOCAT processes shotgun metagenomics Illumina sequencing data to perform quality control, assembly, gene prediction, and taxonomic and functional profiling of microbial communities.

Key Features:

• Modular architecture: Modular architecture permitting customization and interchange of processing components.
• Raw data handling: Read preprocessing of raw fastQ files including quality control to filter low-quality reads and contaminants.
• Assembly: Assembly of metagenomic reads into contigs using assembly programs optimized for high-throughput sequencing.
• Gene prediction: Prediction of protein-coding genes on assembled contigs using advanced gene prediction algorithms.
• Taxonomic and functional profiling: Taxonomic and functional profiling by mapping reads against reference databases, enabling identification of phylogenetic marker genes, extraction or removal of specific sequences, and abundance calculations.
• MOCAT2 reference catalogs: Automated generation and annotation of non-redundant reference catalogs built from pre-computed assignments across 18 databases covering diverse functional categories (MOCAT2).
• HPC integration: Integration with LSF, PBS, and SGE queuing systems for high-performance computing on UNIX-based platforms.
• Implementation: Implemented in Perl 5 and Python 2.7.
• Output formats: Produces multi-sheet Excel documents and queryable SQL databases summarizing statistics from each processing step.
• License: Distributed under the GPL3 open-source license.
• Benchmarking: Benchmarked on artificial, real, and simulated metagenomes with reported improvements in selected quality metrics.

Scientific Applications:

• Shotgun metagenomics: Processing and analysis of shotgun metagenomics datasets generated with Illumina sequencing to profile microbial community composition and function.
• Gene catalog construction: Assembly of reads and prediction of protein-coding genes to construct gene catalogs and reference databases.
• Functional annotation: Functional annotation and characterization of metagenomes using non-redundant catalogs annotated across 18 functional databases.
• Taxonomic profiling: Generation of taxonomic profiles via identification of phylogenetic marker genes and read mapping.
• Abundance estimation and sequence filtering: Abundance estimation of taxa and genes and removal or extraction of specific sequences.
• Method benchmarking: Benchmarking and validation of metagenomic workflows using artificial, real, and simulated metagenomes.

Methodology:

Computational steps explicitly include raw fastQ quality control and contaminant filtering, assembly of reads into contigs with assembly programs, protein-coding gene prediction on contigs, read mapping to reference databases for taxonomic and functional profiling (including abundance calculations and sequence extraction/removal), automated creation and annotation of non-redundant reference catalogs from pre-computed assignments across 18 databases (MOCAT2), and execution via LSF, PBS, or SGE on UNIX; implemented in Perl 5 and Python 2.7 with outputs as multi-sheet Excel and SQL databases.