V-Phaser

V-Phaser 2 calls and phases low-frequency intrahost variants from ultra-deep paired-end next-generation sequencing data to characterize viral population diversity.

Key Features:

• Paired-end phasing: Utilizes paired-end read data to call phased variants and improve detection accuracy within viral genomes.
• Length polymorphism inference: Represents and infers length polymorphisms to capture insertion/deletion variation in viral populations.
• Error filtering: Implements an in-line filter to remove erroneous calls arising from systematic sequencing artifacts.
• Covariation-based sensitivity: Leverages covariation information between observed variants to enhance sensitivity for low-frequency variants.
• Expectation-Maximization recalibration: Employs an expectation-maximization algorithm to iteratively recalibrate base quality scores and increase specificity.
• Platform support: Handles data from next-generation sequencing platforms including Illumina and 454 Titanium.
• Performance optimization: Optimized for memory and runtime to process extremely deep paired-end sequencing data efficiently.

Scientific Applications:

• Intrahost variant detection: Detection and phasing of low-frequency intrahost variants from ultra-deep sequencing (tens to hundreds of thousands-fold coverage).
• Viral population diversity and evolution: Characterization of genetic diversity and evolutionary dynamics within viral populations.

Methodology:

Combines paired-end read phasing, representation and inference of length polymorphisms, covariation analysis between variants, an expectation-maximization algorithm for iterative base quality score recalibration, and an in-line filter to remove systematic sequencing artifacts; supports Illumina and 454 Titanium data and includes memory/runtime optimizations.