QNet

QNet constructs split networks from weighted quartet trees to analyze phylogenetic relationships using an agglomerative, quartet-based alternative to the distance-based Neighbor-Net (NNet) method.

Key Features:

• Quartet-based construction: Builds split networks directly from weighted quartet trees and quartet data.
• Agglomerative computation: Agglomeratively computes a collection of circular weighted splits for a given taxa set.
• Planar split representation: Represents computed circular weighted splits as a planar split network.
• Alternative to distance methods: Operates on character-based quartet data when distance matrices are unavailable, providing an alternative to distance-based techniques.
• Methodological counterpart: Serves as a quartet-based counterpart to the distance-based Neighbor-Net (NNet) algorithm.
• Analytical assessment: Assesses the appropriateness of quartet-based tree-building methods for specific datasets.
• Implementation: Implemented in Java.

Scientific Applications:

• Phylogenetic network inference: Construction of phylogenetic split networks from weighted quartet data.
• Method evaluation: Evaluating whether datasets support quartet-based tree-building approaches.
• Comparative analysis: Providing an alternative framework to Neighbor-Net for analyses that prioritize character-based information over distances.
• Empirical application: Applied to a published Salmonella dataset.

Methodology:

Agglomerative aggregation of weighted quartet trees to compute circular weighted splits for taxa, followed by representation of those splits as a planar split network.