Caterpillars on three and four leaves are sufficient to reconstruct binary normal networks

Area of Science:

• Computational Biology
• Phylogenetics
• Network Theory

Background:

• Rooted binary phylogenetic trees are determined by displayed rooted triples.
• This does not hold for arbitrary rooted binary phylogenetic networks, leading to non-isomorphic networks with identical triple sets.
• Previous research showed limitations in reconstructing networks from small subnetworks like binets and trinets.

Purpose of the Study:

• To investigate if specific subtrees can uniquely determine rooted binary normal networks.
• To establish a constructive method for reconstructing these networks.

Main Methods:

• Analyzing displayed rooted triples and their limitations for network determination.
• Focusing on specific subtrees: caterpillars on three and four leaves.
• Developing a constructive proof leading to a polynomial-time algorithm.

Main Results:

• Rooted binary normal networks are uniquely determined by their sets of displayed caterpillars on three and four leaves.
• A polynomial-time algorithm is presented for reconstructing such networks.

Conclusions:

• Caterpillars on three and four leaves serve as sufficient information to uniquely identify rooted binary normal networks.
• This provides a novel and efficient method for phylogenetic network reconstruction.