ELM: enhanced lowest common ancestor based method for detecting a pathogenic virus from a large sequence dataset.

Area of Science:

• Bioinformatics
• Virology
• Computational Biology

Background:

• Emerging viral diseases pose a significant public health threat, often originating from animal-to-human transmission.
• Effective surveillance for pathogenic viruses is crucial for preparedness against potential outbreaks.
• Next-generation sequencing (NGS) enables virus identification without specific primers, but taxonomic identification of vast sequence data is computationally intensive.

Purpose of the Study:

• To develop an efficient method for identifying viruses from massive sequence data.
• To overcome the computational bottleneck associated with traditional BLAST searches in NGS analysis.

Main Methods:

• Proposed an enhanced lowest-common-ancestor (ELM) based method for virus identification.
• Utilized a customized database of viral sequences to reduce computational cost.
• Implemented a novel criterion to minimize false positive assignments from the specialized database.

Main Results:

• Achieved virus identification over 1,000 times faster than conventional methods.
• Maintained accuracy comparable to traditional approaches.
• Demonstrated the effectiveness of ELM in handling massive sequence data.

Conclusions:

• ELM is anticipated to aid in the direct diagnosis of viral infections.
• The ELM web server and viral database are publicly accessible for research use.