Unveiling pathogens and contaminants: refining metagenomics for clinical diagnostics

Area of Science:

• Clinical diagnostics
• Genomics
• Infectious disease research

Background:

• Shotgun metagenomic sequencing (mNGS) is a powerful untargeted method for pathogen detection and genome characterization.
• Clinical implementation of mNGS is hindered by contamination and reduced sensitivity, particularly in low-biomass samples.

Purpose of the Study:

• To assess the performance of mNGS in clinical samples for pathogen detection and genome characterization.
• To develop and evaluate a framework for improving contamination management and sensitivity in clinical mNGS.

Main Methods:

• Applied mNGS to 144 clinical samples (chronic, acute, and respiratory co-infections).
• Established a contamination control framework using negative controls, contaminant watchlists, and computational filtering.
• Assessed viral detection and genome recovery across different sample types and viral loads.

Main Results:

• Viral load was the primary determinant of mNGS sensitivity; reliable recovery occurred at higher titers.
• The developed framework significantly improved contamination management, reducing false positives and enhancing viral genome recovery.
• mNGS detected clinically relevant co-infections and refined viral classification, but highlighted risks of spurious detections without proper workflows.

Conclusions:

• Defined practical sensitivity thresholds for clinical mNGS.
• Emphasized the critical need for contamination-aware workflows, especially for low-biomass samples.
• Provided an open-source contaminant watchlist to improve the reliability and utility of clinical metagenomics.