MinPath

MinPath reconstructs a minimal set of biological pathways from protein family predictions using a parsimony-based algorithm to reduce false-positive pathway assignments in genomic and metagenomic datasets.

Key Features:

• Parsimony-based reconstruction: Identifies a minimal set of pathways that together explain the observed protein families using a parsimony criterion.
• Protein family input: Uses protein family predictions derived from query sequences, explicitly including KEGG Orthology (KO) families and FIGfams.
• Pathway mapping: Maps detected protein families to known pathway definitions from resources such as KEGG and SEED.
• Spurious annotation filtering: Filters out spurious pathway annotations by requiring essential biochemical steps for pathway functionality.
• False-positive reduction: Produces a reduced set of inferred pathways compared to direct mapping approaches, lowering pathway overestimation.
• Genome and metagenome application: Applicable to individual genomes (e.g., KEGG genomes) and metagenomic datasets to refine pathway assignments.

Scientific Applications:

• Genome pathway annotation: Provides conservative pathway inference for genome annotations using KO and FIGfams mappings.
• Metagenomic functional profiling: Refines estimates of pathways encoded by microbial communities in metagenomes by reducing false positives.
• Comparative functional diversity analysis: Enables more accurate comparisons of pathway repertoires across samples by minimizing spurious pathway calls.
• Validation of pathway assignments: Identifies and eliminates incorrect pathway assignments in reference genome datasets such as KEGG genomes.

Methodology:

Identify protein families (e.g., KO, FIGfams) from query sequences, map those families to known pathway definitions, and apply a parsimony algorithm to select the minimal set of pathways while filtering annotations by required biochemical steps.