HMMPolish: a coding region polishing tool for TGS-sequenced RNA viruses

Area of Science:

• Genomics
• Bioinformatics
• Virology

Background:

• Accurate viral genomes are crucial for downstream analyses like lineage identification and variant monitoring.
• Third-generation sequencing (TGS) offers long reads for virus sequencing but has a higher error rate than next-generation sequencing.
• Existing polishing tools show promise but have limitations in correcting errors, especially within coding regions.

Purpose of the Study:

• To develop a novel pipeline, HMMPolish, for correcting errors in protein-coding regions of known RNA viruses.
• To improve the accuracy of viral genome assembly from TGS data.
• To address the limitations of current polishing tools in handling errors within critical genomic regions.

Main Methods:

• Developed HMMPolish, a pipeline utilizing profile Hidden Markov Models (HMMs) of protein families/domains.
• Applied HMMPolish to raw TGS reads and assembled sequences.
• Utilized HMMs to identify and correct errors missed by other polishers, focusing on protein-coding regions.

Main Results:

• HMMPolish demonstrated effective error correction in protein-coding regions of RNA viruses.
• Validated on 34 datasets covering HIV-1, influenza-A, norovirus, and SARS-CoV-2.
• Outperformed or competed favorably with popular polishing tools in benchmark tests.

Conclusions:

• HMMPolish offers a significant improvement for correcting errors in viral protein-coding regions.
• The tool enhances the accuracy of viral genome assembly from TGS data.
• HMMPolish is valuable for accurate lineage identification and variant monitoring in RNA viruses.