PanFP

PanFP infers the functional composition of microbial communities from 16S rRNA gene survey data by constructing pangenome-based functional profiles for operational taxonomic units (OTUs).

Key Features:

• Pangenome Reconstruction: PanFP constructs a pangenome for each operational taxonomic unit (OTU) by aggregating known genes and genomes from the OTU's taxonomic lineage.
• Functional Profile Assignment: It assigns functional profiles to OTUs by weighting the reconstructed pangenome's functional profile by the observed abundance of each OTU in a sample.
• Flexibility in OTU Building: The method accepts any OTU building approach and is not restricted to closed-reference OTU picking or specific reference databases.
• Validation and Performance: Performance was validated by comparison to shotgun metagenomic measurements across 65 diverse communities, reporting Spearman correlation coefficients of 0.8–0.9 for deeply sequenced samples and showing comparable performance to PICRUSt.
• Cost-Effectiveness: By leveraging 16S rRNA gene survey data instead of shotgun metagenomics, PanFP enables lower-cost inference of community functional profiles for comparative analyses.

Scientific Applications:

• Ecological and Environmental Microbiology: Inferring community functional composition in ecological and environmental microbiology studies where shotgun metagenomics is not feasible.
• Comparative Functional Metagenomics: Enabling comparative functional metagenomic and metadata analyses across large-scale 16S rRNA datasets.
• Diversity–Function Relationship Studies: Exploring microbial diversity–function relationships across diverse environments using 16S rRNA surveys.

Methodology:

PanFP constructs pangenomes per OTU by aggregating known genes and genomes from taxonomic lineages, assigns functions by weighting pangenome functional profiles with observed OTU abundances, accepts any OTU-building method (not restricted to closed-reference), and was validated by comparing inferred profiles to shotgun metagenomic measurements across 65 communities using Spearman correlation.