Phylogenetic network analysis as a parsimony optimization problem.

Area of Science:

• Phylogenetics
• Comparative Biology
• Evolutionary Biology

Background:

• Phylogenetic networks, which allow multiple ancestors, are increasingly used in comparative biology over traditional trees.
• Existing network types (softwired and hardwired) have limitations in biological interpretation and hypothesis testing.
• Current methods prevent simultaneous hypothesis testing between trees and networks due to differing cost calculations.

Purpose of the Study:

• To propose a network cost adjustment enabling equal footing for phylogenetic trees and soft-wired networks in parsimony-based analyses.
• To facilitate simultaneous hypothesis testing among trees and networks.

Main Methods:

• Introduced a cost adjustment (penalty) for soft-wired phylogenetic networks.
• Applied this cost adjustment to real and simulated biological datasets.

Main Results:

• The proposed cost adjustment allows phylogenetic trees and soft-wired networks to compete on an equal parsimony optimality basis.
• Evaluations on diverse datasets showed that the favored representation (tree or network) aligned with expectations or simulation parameters.

Conclusions:

• The developed soft-wired network cost regime provides a quantitative criterion for optimality-based searches.
• This approach enables trees and networks to participate in hypothesis testing concurrently, advancing phylogenetic comparative methods.