PathoLive-Real-Time Pathogen Identification from Metagenomic Illumina Datasets

Area of Science:

• Microbiology
• Bioinformatics
• Computational Biology

Background:

• Next-generation sequencing (NGS) is vital for pathogen diagnostics but suffers from long turnaround times due to post-sequencing analysis.
• Interpreting NGS data is complicated by contaminations, irrelevant sequences, and data complexity.
• Current methods delay pathogen identification until after sequencing is complete.

Purpose of the Study:

• To develop a real-time diagnostics pipeline (PathoLive) for rapid pathogen detection from clinical samples.
• To enable pathogen identification hours before sequencing completion.
• To provide accurate prioritization and visualization of pathogen hits based on clinical significance.

Main Methods:

• Implemented PathoLive, a real-time bioinformatics pipeline for pathogen detection.
• Utilized real-time alignment with HiLive2 for sequence mapping.
• Scored sequence mappings against common contaminants, low-entropy regions, and non-pathogenic organisms.
• Developed an interactive taxonomic tree for visualizing and interpreting results.

Main Results:

• PathoLive detected all six spiked pathogenic viruses in a human plasma sample after only 40 of 200 sequencing cycles.
• Correctly identified Crimean-Congo hemorrhagic fever virus in a real-world sample from Sudan.
• Detected SARS coronavirus in a 2019 Wuhan dataset, even without the novel virus reference genome.
• Effectively de-emphasized clinically irrelevant hits across all tested samples.

Conclusions:

• PathoLive provides fast and accurate NGS-based pathogen identification.
• The pipeline successfully prioritizes and visualizes pathogen hits by clinical significance.
• PathoLive is an open-source tool available on GitLab and BioConda, facilitating broader adoption.