TreeKnit

TreeKnit infers ancestral reassortment graphs (ARGs) from segment phylogenies to identify reassortment events in segmented genomes such as human influenza.

Key Features:

• ARG inference from two segment trees: Identifies ancestral reassortment graphs by comparing the evolutionary histories of two genome segment trees.
• Topological comparison: Leverages topological differences between segment trees to locate regions compatible with reassortment events.
• Greedy algorithm: Applies a greedy algorithmic approach to efficiently identify likely reassortment regions.
• Validation on simulated genealogies: Validated using simulated genealogies with known reassortment events to assess accuracy.
• Comparative performance: Demonstrates accuracy matching more complex Bayesian methods while providing orders-of-magnitude faster runtime.
• Resolution enhancement: Improves the resolution of segment trees by incorporating inferred reassortment events.
• Reassortment visualizations: Produces information suitable for constructing visualizations that depict reassortment events.
• Implementation: Provided as a Julia package.

Scientific Applications:

• Influenza reassortment analysis: Elucidates reassortment patterns and impacts in seasonal human influenza viruses.
• Evolutionary studies of segmented viruses: Supports investigations into the role of reassortment in viral evolution and genetic diversity.
• Surveillance and outbreak investigation: Aids monitoring and interpretation of reassortment-driven emergence of novel viral strains relevant to outbreak risk assessment.

Methodology:

Infers ARGs from two segment trees by comparing tree topologies and applying a greedy algorithm, with validation on simulated genealogies and performance compared to Bayesian approaches.