PSORTm

PSORTm predicts protein subcellular localization (SCL) from metagenomic sequences by integrating PSORTb components with automated cell envelope classification to enable read-based SCL prediction across diverse microbial communities.

Key Features:

• PSORTb integration: Reuses core SCL prediction components from PSORTb (including PSORTb 3.0.2 framework elements) for metagenomic sequence analysis.
• Metagenomic compatibility: Processes metagenomic sequences directly without requiring assembled genomes, enabling analysis of unassembled reads.
• Automated cell envelope classification: Classifies proteins by cell envelope type automatically to inform SCL prediction across diverse microbial taxa.
• High precision and sensitivity: Demonstrates retained high precision and fragment-length-dependent increased sensitivity based on evaluation results.
• Read-based analysis versus MAGs: Provides read-based SCL prediction that parallels analysis from metagenome-assembled genomes (MAGs) while eliminating the need for manual MAG classification.

Scientific Applications:

• Human microbiome studies: Identifies subcellularly localized proteins within human-associated microbial communities.
• Agri-food organism research: Profiles protein localization in microorganisms relevant to agriculture and food systems.
• Environmental microbiology: Analyzes environmental metagenomes, including freshwater samples, to identify localization-linked markers.
• Biomarker discovery for diagnostics: Aids identification of cell-surface biomarkers for rapid diagnostic test development.
• Water quality monitoring: Supports discovery of potential markers of water quality from freshwater metagenomic data.
• Functional profiling of microbial communities: Facilitates uncovering functional insights relevant to medical diagnostics, environmental monitoring, and industrial applications.

Methodology:

PSORTm combines PSORTb SCL prediction modules with an automated cell envelope classification system and was evaluated using 5‑fold cross-validation on in silico‑fragmented sequences, showing retained precision and sensitivity that increases with fragment length.