PhyCovA

PhyCovA facilitates exploration of covariates influencing pathogen spread by parameterizing discrete phylogeographic models as generalized linear models and performing regression analyses on phylogenetic trees annotated with discrete states.

Key Features:

• Rapid Exploration: Enables swift exploration of associations between candidate covariates and phylogenetically informed transition events across discrete geographical locations, relevant for analyses of viral genomes during outbreaks.
• Input Flexibility: Accepts phylogenetic trees with discrete state annotations at internal nodes and performs ancestral state reconstruction when such annotations are missing.
• Generalized Linear Model Parameterization: Supports parameterizing discrete phylogeographic models as generalized linear models to identify predictors of viral lineage dispersal among discrete locations.
• Regression Analyses: Performs univariate and multivariate linear regression analyses to assess the impact of covariates on pathogen dispersal.
• Exploratory Variable Selection: Provides exploratory variable selection analysis to identify significant predictors from a set of candidate covariates.
• Visualization: Produces visualizations for regression results and for phylogenetic trees annotated by ancestral state reconstruction.

Scientific Applications:

• Phylogeographic Covariate Analysis: Identifies covariates associated with viral lineage movements among discrete locations using phylogenetically informed transition events.
• Epidemic Outbreak Investigation: Facilitates rapid assessment of factors influencing pathogen dispersal during epidemic outbreaks using viral genome data.
• Public Health Inference: Supports identification of predictors that can inform public health strategies and interventions targeting disease transmission.

Methodology:

Parameterizes discrete phylogeographic models as generalized linear models and uses phylogenetic trees annotated with discrete state information (with ancestral state reconstruction when needed) to perform univariate and multivariate linear regression and exploratory variable selection to associate covariates with location-to-location transition events.