VERSO

VERSO processes variants profiles from viral samples to produce phylogenetic models of viral evolution.

Key Features:

• Boolean Matrix Factorization: Solves a Boolean Matrix Factorization problem with phylogenetic constraints by maximizing a log-likelihood function.
• Two-step framework: Implements VERSO STEP 1 as an R package, with STEP 2 as a separate subsequent step.
• Variant profile processing: Processes mutation profiles from viral genome sequencing samples.

Scientific Applications:

• Viral evolution tracking: Tracks virus evolution through mutations that accumulate in viral genomes during a pandemic.
• Infection network reconstruction: Unravels viral infection networks from genomic data.
• Cryptic transmission detection: Uncovers cryptic transmissions to estimate the proportion of undetected infected people.
• Epidemiological trend prediction: Predicts likely trends in infected, hospitalized, dead and recovered people.
• Genomic epidemiology modeling: Estimates models from genomic epidemiology via viral sequencing samples.

Methodology:

VERSO solves a Boolean Matrix Factorization problem with phylogenetic constraints by maximizing a log-likelihood function. The framework consists of two separate and subsequent steps; this package provides an R implementation of VERSO STEP 1.