MGRAST

MG-RAST, short for Metagenomics RAST, is a web-based platform for the analysis of metagenomic sequence data. It provides automated pipelines that process raw reads through quality control, gene prediction, annotation, and the generation of taxonomic and functional profiles. The service is designed to support large-scale comparative metagenomics studies by enabling analyses across multiple samples and datasets.

Central to MG-RAST is the M5nr annotation framework, which integrates annotations from multiple public databases to assign taxonomy and function to predicted genes. Outputs typically include per-sample taxonomic abundances, functional category distributions, and cross-sample comparisons. Users can browse results online or download tables and visualization-ready files for downstream analysis.

Data submission is submission-friendly: users can upload FASTA or FASTQ files or provide links to data, and projects can be kept private or released publicly. MG-RAST also stores associated metadata such as sample origin and experimental details. The platform supports batch submissions and offers several analysis options and parameter settings to tailor the pipeline to the data.

Typical use cases include characterizing microbial community composition, inferring potential metabolic capabilities, and performing cross-study comparisons. Limitations include the dependence on reference databases for annotation, potential biases introduced by assembly-free approaches, and varying computational requirements. MG-RAST remains a widely used resource in metagenomics research for its accessibility and standardized workflows.