Trinets encode tree-child and level-2 phylogenetic networks.

Area of Science:

• Evolutionary biology
• Computational phylogenetics
• Network theory

Background:

• Phylogenetic networks model complex evolutionary histories beyond simple trees, incorporating reticulate processes like hybridization and gene transfer.
• Constructing phylogenetic networks from simpler data, such as rooted triplets (trees on three species), is an active area of research.
• Rooted triplets do not always uniquely encode phylogenetic networks, necessitating extensions like trinets.

Purpose of the Study:

• To investigate whether trinets, a generalization of rooted triplets, can encode specific classes of rooted phylogenetic networks.
• To provide theoretical support for the conjecture that trinets encode all "recoverable" rooted phylogenetic networks.
• To explore potential new methods for constructing phylogenetic networks based on trinet decompositions.

Main Methods:

• Proving two decomposition theorems based on trinets.
• Applying these theorems to binary level-2 networks and binary tree-child networks.
• Analyzing the encoding properties of trinets for these network classes.

Main Results:

• Demonstrated that trinets encode recoverable binary level-2 phylogenetic networks.
• Showed that trinets encode binary tree-child phylogenetic networks.
• Established decomposition theorems for recoverable binary rooted phylogenetic networks based on trinets.

Conclusions:

• The study provides evidence supporting the conjecture that trinets encode all recoverable rooted phylogenetic networks.
• The findings suggest that trinets are a powerful tool for understanding and potentially constructing complex phylogenetic networks.
• The decomposition theorems could pave the way for novel algorithms in phylogenetic network inference.