Dollo Parsimony Overestimates Ancestral Gene Content Reconstructions

Area of Science:

• Evolutionary biology
• Genomics
• Bioinformatics

Background:

• Ancestral reconstruction is key to understanding gene family evolution.
• Dollo parsimony is a common but potentially flawed method for this task.
• Its assumption that characters are not regained conflicts with molecular evolution realities.

Purpose of the Study:

• To test the suitability of Dollo parsimony for ancestral gene content reconstruction.
• To compare Dollo parsimony with a maximum likelihood approach.
• To evaluate method performance using simulated and real eukaryotic data.

Main Methods:

• Simulated 5,000-gene datasets with varying evolutionary rates.
• Applied Dollo parsimony and maximum likelihood methods for ancestral reconstruction.
• Reconstructed protein domain evolution on a eukaryotic phylogeny.

Main Results:

• Dollo parsimony frequently overestimated ancestral gene content.
• Overestimations were most pronounced near the root of the phylogenetic tree.
• Maximum likelihood methods provided a more accurate reconstruction.

Conclusions:

• Dollo parsimony is inappropriate for ancestral reconstruction based on sequence homology.
• Its assumption of non-reciprocal gain hinders accurate evolutionary inference.
• Maximum likelihood approaches are recommended for gene family evolution studies.