QuasiSeq

QuasiSeq profiles viral quasispecies from high-throughput sequencing data by transforming identification of rare variants in PacBio long reads into a clustering problem to accommodate high sequence similarity and long-read error rates.

Key Features:

• SigClust: Employs a signature-based self-tuning clustering method (SigClust) to enhance accuracy and sensitivity in profiling viral mixtures.
• Clustering transformation: Reformulates quasispecies identification as an efficient clustering problem for closely related sequences.
• Long-read support: Processes high-throughput PacBio long reads and accounts for long-read sequencing error rates.
• Low-quality read handling: Identifies quasispecies from sequencing reads with accuracy below 80%.
• High-accuracy with circular consensus reads: Achieves 100% accuracy when using high-quality circular consensus sequencing reads.
• Scalability and parallelization: Allows adjustment of signature size to control computational resources and implements parallel computation to expedite cluster processing.

Scientific Applications:

• Viral quasispecies profiling: Generates intra-host viral diversity profiles from sequencing data.
• Viral evolution studies: Supports analysis of mutation dynamics and viral evolution over time.
• Epidemiology and surveillance: Enables tracking of viral variants for epidemiological investigations.
• Antiviral resistance detection: Aids detection and monitoring of antiviral resistance-associated variants.

Methodology:

Transforms quasispecies identification into a clustering problem and applies the signature-based self-tuning clustering method SigClust to PacBio long reads (including circular consensus sequencing reads), with adjustable signature sizes and parallel computation to expedite cluster processing.