Predicting microbial traits with phylogenies.

Area of Science:

• Microbial Ecology
• Evolutionary Biology
• Bioinformatics

Background:

• Phylogenetic conservatism, the tendency of related species to share traits, is widely applied in microbial ecology to predict traits like gene content or environmental tolerance.
• Microbial ecologists face challenges including large phylogenies and limited trait data, potentially compromising trait reconstruction accuracy.

Purpose of the Study:

• To review the prevalence and strength of phylogenetic trait conservatism across bacterial and archaeal traits.
• To assess the accuracy of phylogeny-based trait prediction methods under common microbial ecology scenarios.
• To provide recommendations for using phylogeny-based trait predictions in microbial research.

Main Methods:

• Literature review of 990 bacterial and archaeal traits.
• Analysis of phylogenetic trait conservatism.
• Simulation exercises to evaluate trait prediction accuracy based on phylogeny size, trait conservatism, and data availability.

Main Results:

• Phylogenetic trait conservatism is widespread but generally weak for ecologically relevant phenotypic traits, while strong for genetically complex traits.
• Approximately 60% of the variation in phylogeny-based trait predictions is influenced by trait conservatism, tree size, data completeness, and proximity to known trait values.
• Prediction accuracy is comparable for both binary and continuous traits.

Conclusions:

• Phylogenetic trait conservatism is a common but variable phenomenon in microbes, with implications for trait prediction accuracy.
• Microbial ecologists must carefully consider factors like trait conservatism and data availability when using phylogeny-based predictions.
• Recommendations are provided for optimizing the application of phylogeny-based trait predictions in microbial ecology studies.