MEGAN

MEGAN analyzes metagenomic DNA sequences to assign reads or contigs to taxa using sequence comparisons and a lowest common ancestor algorithm and to summarize the taxonomic composition of microbial communities.

Key Features:

• Preprocessing with Sequence Comparison: Compares DNA reads or contigs against established sequence databases using BLAST or alternative sequence comparison tools.
• Taxonomical Analysis and Summarization: Computes and summarizes taxonomic content using the NCBI taxonomy to organize results.
• Lowest Common Ancestor Algorithm: Assigns reads to taxa via a lowest common ancestor algorithm that reflects sequence conservation and does not require assembly or targeting specific phylogenetic markers.
• Graphical and Statistical Output: Produces graphical and statistical summaries for comparing metagenomic datasets.
• Application to Diverse Data Sets: Has been applied to datasets including the Sargasso Sea dataset, a metagenomic sample from a mammoth bone, and several complete microbial genomes.
• Performance Evaluation through Simulations: Performance has been evaluated using simulations that consider different read lengths to assess effects of sequencing technology and read length on assignment.

Scientific Applications:

• Taxonomic profiling: Derives taxonomic composition and diversity of microbial communities from metagenomic datasets.
• Comparative metagenomics: Enables comparison of taxonomic profiles across different datasets to identify differences in community composition.
• Environmental and ecological studies: Supports environmental microbiology and ecology research by analyzing community composition from environmental samples such as the Sargasso Sea.
• Evolutionary biology: Facilitates study of microbial diversity and phylogenetic patterns without requiring marker-gene targeting or assembly.
• Assessment of sequencing strategies: Assesses the impact of read length and sequencing technology on taxonomic assignment through simulation-based evaluation.

Methodology:

Computational steps include sequence comparison of reads or contigs to reference databases using BLAST or alternatives, taxonomic summarization via the NCBI taxonomy, read-to-taxon assignment using a lowest common ancestor algorithm, and simulation-based performance evaluation across different read lengths.