MEBS

MEBS computes an entropy-based score to evaluate and infer metabolic capabilities, particularly sulfur-cycle metabolic machinery, from large genomic and metagenomic datasets.

Key Features:

• Metabolic Pathway Evaluation: Assesses and compares metabolic processes including biogeochemical cycles such as the sulfur cycle.
• Entropy-Based Scoring System: Uses relative entropy (H΄) to quantify enrichment of Pfam protein domains and construct an entropy-based score indicating presence of target metabolic machinery.
• Domain and Genome Curation: Curates compounds, genes, pathways, microbial taxa, characteristic Pfam protein domains, and complete-sequenced genomes relevant to the sulfur cycle into representative datasets.
• Marker Domain Identification: Identifies marker Pfam domains within metagenomic sequences, exemplified by DsrC (PF04358), that are indicative of specific metabolic functions.

Scientific Applications:

• Biogeochemical cycle modeling: Models complex cycles, with emphasis on the sulfur cycle, by scoring presence of pathway components across genomes and metagenomes.
• Genome classification: Classifies genomes from hard-to-culture organisms, including Candidatus Desulforudis audaxviator, using curated Pfam domain signatures.
• Environmental metagenomics analysis: Detects sulfur-related metabolic capabilities in metagenomes from environments such as hydrothermal vents and deep-sea sediments.

Methodology:

Data collection from non-redundant microbial genomes in RefSeq and metagenomes from MG-RAST; calculation of relative entropy to assess Pfam domain enrichment in sulfur-related genomes and construct the score; performance evaluation via random sampling, linear regression models, receiver operator characteristic (ROC) plots and area under the curve (AUC) metrics (AUC = 0.985).