VarSel

R implementation of the environment-dependent birth-death model within a Bayesian framework.

Functions

Directory Functions contains the main functions that are used in the analyses:

• halfcauchy: compute density and random deviates of an half-Cauchy distribution.
• halfnormal: compute density and random deviates of an half-normal distribution.
• laplace: compute density and random deviates of a Laplacian distribution.
• likelihood_bd_mod_c: compute the likelihood of a user-defined environment-dependent birth-death process.
• MLE: compute the maximum likelihood estimates of the parameters of a user-defined environment-dependent birth-death process.
• sim_env_bd_mod_c: generate a random phylogenetic tree under a user-defined environment-dependent birth-death process.
• run_env_bd_MCMC: generate three parallel independent MCMC chains sampled from the posterior of a user-defined environment-dependent birth-death process, using user defined uninformative priors and MH sampling.
• proposal: proposal functions for the MH-MCMC.

Data

Directory Data constains the environmental variables used in the analyses.

Phylo

Directory phylo constain the phylogenetic trees used in the analyses.

Treesim_onevar

Example of how simulating a phylogenetic tree using only one environmental dependency.

Rscript Treesim_onevar.R [env_var] [seed] [extinction_rate]

Options for env_var: Temperature, CO2, SeaLevel, d13C, Silica. Options for extinction_rate: const, ratio.

Treesim_onevar

Example of how simulating a phylogenetic tree using two environmental dependencies.

Rscript Treesim_onevar.R [env_var_1] [env_var_2] [seed] [extinction_rate]

Options for env_var_1 and env_var_2: Temperature, CO2, SeaLevel, d13C, Silica. Options for extinction_rate: const, ratio.

Bayes_onevar

Example of how performing a MH-MCMC sampling of the posterior with only one environmental dependency.

Rscript Bayes_onevar.R [seed] [prior] [extinction_rate]

Options for prior: unif, norm, exp. Options for extinction_rate: const, ratio.

Bayes_twovar

Example of how performing a MH-MCMC sampling of the posterior with two environmental dependencies.

Rscript Bayes_onevar.R [seed] [prior] [extinction_rate]

Options for prior: unif, norm, exp. Options for extinction_rate: const, ratio.

Variable_selection_onevar

Example of how performing a MH-MCMC sampling of the horseshoe posterior implementing the variable selection procedure.

Rscript Variable_selection_onevar.R [seed]