ViQUF

ViQUF reconstructs de novo viral quasispecies from short-read sequencing data to assemble haplotypes and estimate their frequencies for analysis of intrahost mutation and recombination.

Key Features:

• De Novo Assembly: Performs assembly without reliance on reference genomes to resolve multiple haplotypes from short-read sequencing data.
• de Bruijn Graph Construction: Builds an initial assembly graph using a de Bruijn graph approach for scalable graph representation.
• Paired-End Integration: Incorporates paired-end information between adjacent vertices to inform connections in the assembly graph.
• Min-Cost Flow Network: Constructs a flow network and applies a min-cost flow algorithm to produce an approximate paired assembly graph with edge labels representing frequency estimations.
• Greedy Path Reconstruction: Extracts haplotypes by performing greedy path reconstruction guided by the min-cost flow solution on the approximate paired assembly graph.
• Outputs and Performance: Outputs contigs with frequency estimates and reports performance at least four times faster than previous methods while using up to half the memory, maintaining or surpassing assembly and frequency-estimation quality of overlap graph-based methods.

Scientific Applications:

• Viral population diversity analysis: Reconstructs intra-host viral haplotypes to characterize quasispecies diversity arising from mutation and recombination.
• Mutation rate and evolution studies: Provides haplotype-resolved data useful for inferring mutation dynamics and evolutionary trajectories.
• Drug resistance and vaccine research: Identifies variant haplotypes relevant to drug resistance mechanisms and vaccine-targeted antigenic variation.
• Benchmarking on datasets: Applies to both real-world and simulated sequencing datasets for viral genomics and epidemiology studies.

Methodology:

Constructs a de Bruijn assembly graph, integrates paired-end links between adjacent vertices, builds a flow network and solves a min-cost flow problem to obtain an approximate paired assembly graph with frequency-labeled edges, and performs greedy path reconstruction guided by the min-cost flow solution to produce contigs with frequency estimates.