EmpPrior

EmpPrior facilitates incorporation of external phylogenetic datasets from TreeBASE to specify informative branch-length priors for Bayesian phylogenetic analyses, improving the accuracy and precision of tree-length and posterior estimates.

Key Features:

• Data-driven prior specification: Systematically searches TreeBASE to identify datasets with similar taxonomic and genetic sampling and uses them to inform branch-length prior settings.
• Two-component architecture: EmpPrior-search (Java) queries TreeBASE to retrieve relevant datasets and EmpPrior-fit (R) parameterizes branch-length distributions from those retrieved data.
• Improved estimation accuracy: Incorporating external data into branch-length priors increases the accuracy and precision of tree-length estimates in Bayesian phylogenetic models.
• Extensible priorization strategy: The approach for parameterizing priors based on external datasets can be adapted to other prior parameterization problems in phylogenetics.

Scientific Applications:

• Evolutionary relationship inference: Produces more accurate branch-length estimates to improve phylogenetic reconstructions and topology-based interpretations.
• Divergence-time estimation: Provides empirically informed branch-length priors that can influence posterior estimates of divergence times in Bayesian analyses.
• Lineage diversification studies: Refines branch-length estimates used to infer rates and patterns of lineage diversification.
• Taxon- and gene-sampling-specific analyses: Benefits studies where focal datasets have taxonomic or genetic sampling comparable to datasets available in TreeBASE.

Methodology:

EmpPrior-search queries TreeBASE to identify datasets with similar taxonomic and genetic sampling; EmpPrior-fit in R parameterizes branch-length distributions from the retrieved datasets for use as priors in Bayesian phylogenetic models.