phypanda

phypanda is the Python library for PaNDA — efficient optimization of phylogenetic diversity in phylogenetic networks. It works on directed phylogenetic networks represented in phylozoo, and uses scanwidth (scanwidth) to obtain tree extensions and run fixed-parameter tractable dynamic programs for several budgeted problems.

What you can do here

• Score a chosen set of taxa under different PD notions (all-paths / MAPPD, max-tree, min-tree, and stubs for tree-only, network, and average-tree measures).

• Optimize under an integer budget with taxon costs (notably MAPPD via node-scanwidth FPT, max-tree PD, and greedy / generic wrappers).

• Provide a precomputed tree extension (e.g. from NSW / XP ordering) so repeated solves on the same network avoid rescanning.

The algorithms and objective definitions are those in the PaNDA paper; this site focuses on how to run the code. For definitions and proofs, see the reference below.

Where to read next

User guide

• Installation
  • Requirements
  • Install the package
  • Optional extras
  • Dependencies (automatic with pip install phypanda)
  • Related software
• Quickstart
  • Import pattern
  • Which PD measures are implemented?
  • How a typical solve works
  • High-level helpers
  • Examples
  • Next steps
• API reference
  • Base helpers
  • Measure protocol
  • All-paths (MAPD) measure
  • All-paths solvers
  • Max-tree PD
  • Min-tree PD
  • Tree diversity (stub)
  • Network diversity (stub)
  • Average-tree diversity (stub)

Paper and citation

If you use this software, please cite:

PaNDA: Efficient Optimization of Phylogenetic Diversity in Networks. Niels Holtgrefe, Leo van Iersel, Ruben Meuwese, Yukihiro Murakami, Jannik Schestag. bioRxiv, 2025. https://doi.org/10.1101/2025.11.14.688467

The repository README links the GUI, experiment scripts (under experiments/), and supplementary datasets for the first MAPPD study (experiments/MAPPD/) and budgeted benchmarks (experiments/budgeted_PD/).