AN EMPIRICAL COMPARISON OF MICROCOMPUTER PARSIMONY PROGRAMS, II.

Area of Science:

• Phylogenetics
• Computational Biology
• Systematic Zoology

Background:

• Microcomputer programs for cladistic analysis often suffer from arbitrary resolution and redundancy.
• Previous software had limitations in data set size, cladogram storage, and execution speed.
• Benchmarking earlier mainframe and microcomputer packages is crucial for evaluating new tools.

Purpose of the Study:

• To evaluate the effectiveness and efficiency of J. S. Farris's Hennig86 microcomputer parsimony program.
• To compare Hennig86 performance against established benchmarks and earlier software.
• To assess Hennig86's ability to overcome limitations of existing microcomputer cladistic programs.

Main Methods:

• Evaluation of Hennig86 (version 1.5) using 60 diverse data sets.
• Comparison with data sets used for benchmarking previous mainframe and microcomputer packages.
• Testing exact (exhaustive, minimal) and approximate (branch breaker, mhennig*, tread) analysis options.

Main Results:

• Hennig86 demonstrates superior effectiveness and efficiency compared to earlier parsimony programs.
• The program overcomes arbitrary resolution, redundancy, and limitations in data set size, storage, and speed.
• For approximate analyses, the branch breaker option with mhennig* and tread commands yielded the best performance.

Conclusions:

• Hennig86 represents a substantial advancement in the accuracy and ease of conducting cladistic analyses.
• Earlier parsimony programs, including previous Farris software, are largely rendered obsolete by Hennig86.
• Hennig86 offers robust solutions for phylogenetic analysis on microcomputers.