Novel distances for dollo data.

Area of Science:

• Phylogenetics
• Computational Biology
• Evolutionary Biology

Background:

• Phylogenetic analysis relies on accurate distance metrics for binary data.
• The Dollo model describes trait evolution where traits arise once but can be lost.
• Existing binary distances may not perform optimally under Dollo models.

Purpose of the Study:

• To introduce and evaluate a novel distance metric, the Additive Dollo Distance (ADD), for binary data under a Dollo process.
• To compare the performance of ADD against other established binary distances using simulations.
• To assess the utility of ADD in real-world phylogenetic analyses.

Main Methods:

• Introduction of the Additive Dollo Distance (ADD).
• Simulations of binary data generated under a Dollo model.
• Comparison of ADD with LogDet, restriction-site-based, and other binary distances.
• Application of ADD to Diversity Arrays Technology data and bacterial genome data.

Main Results:

• The Additive Dollo Distance (ADD) demonstrates superior performance on simulated Dollo data compared to other tested distances.
• LogDet distance performed poorly under the Dollo model, suggesting limitations for conditioned genome reconstruction.
• ADD application to Eucalyptus and bacterial gene family data yielded results congruent with previous studies, sometimes with enhanced phylogenetic resolution.

Conclusions:

• The Additive Dollo Distance (ADD) is a theoretically sound and empirically effective metric for phylogenetic analysis of binary data evolving under a Dollo model.
• The findings highlight potential issues with using LogDet distance in specific evolutionary contexts.
• ADD provides a valuable tool for improving phylogenetic resolution in diverse biological datasets.