phylosubsets

phylosubsets generates distributions of time-scaled, species-level phylogenetic trees for extant Mammalia using subset sampling and a backbone-and-patch Bayesian framework to analyze divergence times and diversification rates.

Key Features:

• Subset Sampling: Produces distributions of phylogenetic trees by sampling subsets of taxa to enable focused analyses while retaining broader evolutionary context.
• Credible Sets of Trees: Generates credible sets that capture uncertainty in topology and divergence times across analyses.
• Backbone-and-Patch Approach: Implements a two-level Bayesian inference strategy using a 31-gene supermatrix to estimate major lineage relationships and ages and to build detailed species-level patch phylogenies.
• Backbone Relationships and Ages: Establishes major lineage relationships and node ages using fossil node or tip dating under Bayesian inference.
• Species-Level Patch Phylogenies: Constructs non-overlapping in-group phylogenies (patches) for each representative lineage in the backbone to provide species-level resolution.
• Handling Unsampling of DNA Data: Provides two strategies for taxa lacking DNA sequences: exclude unsampled species in DNA-only trees, or impute missing taxa within taxonomic constraints using branch lengths derived from local birth-death models in completed trees.
• Integration and Comparison: Integrates time-scaled patches with backbone phylogenies to produce comprehensive species-level trees for extant Mammalia and facilitate rate comparisons among lineages.
• Enhanced Accuracy: Improves estimates of recent speciation rates relative to previous supertree approaches that produce unresolved nodes and branch-length artifacts.

Scientific Applications:

• Divergence Time Estimation: Estimating lineage ages and divergence times using fossil node or tip dating within a Bayesian framework.
• Diversification Rate Comparison: Comparing speciation and diversification rates across diverse lineages, including marsupials and placentals, using integrated species-level trees.
• Biodiversity Pattern Analysis: Investigating evolutionary histories and patterns of modern biodiversity across mammalian lineages.
• Uncertainty Quantification in Phylogenetics: Exploring topological and temporal uncertainty via credible sets of trees for downstream comparative analyses.

Methodology:

Uses Bayesian inference on a 31-gene supermatrix with fossil node or tip dating; employs a backbone-and-patch scheme to estimate backbone relationships and ages and to build non-overlapping species-level patch phylogenies; imputes missing taxa within taxonomic constraints using branch lengths from local birth-death models and integrates time-scaled patches with the backbone to produce comprehensive species-level trees.