Skip to contents

InNOutBT

An R package for creating input data and processing output data in formats required of Meade & Pagel’s (2022) BayesTraits computer package.

Overview

The package InNOutBT remains under development. Currently, it has three major functions that help to automate BayesTraits phylogenetic independent contrast regression analyses. To better understand how these functions work, see BayesTraits V4.0.0 Manual for further details on the specific format of the input files that are required by BayesTraits, as well as the BayesTraits strucuture of resulting output files.

Installation

You can install the development version of InNOutBT from GitHub with:

#install.packages("devtools")
devtools::install_github("DBOSlab/InNOutBT")

Usage

A general description of each of the three available functions (phyreg.inputs, phyreg.shell, and phyreg.outputs) are provided below.

1. phyreg.inputs: Creating input mean and linked data for phylogenetic regression analyses

This function produces the required BayesTraits input data of linked samples of trait/climate data (distribution data), by generating two text files, one of which includes the mean values associated with each trait associated with a terminal taxon and a second of which reports samples of trait values that are linked to each other because they were sampled from one accession of the associated terminal taxon. The function also includes an argument to calculate the net node count for each species/accession given a phylogenetic tree. Net node counts have been used to explore the effects of speciation rates on the response variable (e.g. O’Donovan et al. 2018).

Example of using phyreg.inputs: 
library(InNOutBT)
library(treeio)

traitdata <- read.csv("Data/vataireoids_1610_25May2022_BayesTraits.csv")
phylo <- read.beast("Data/vatcomb30_24May22_Yule.con.tree")

phyreg.inputs(tree = phylo,
              data = traitdata,
              tipscol = "terminal",
              NodeCount = TRUE,
              logtransf = c("DBH", "NodeCount"),
              traitcols = c("bio12", "bio15"),
              addtraits = c("log10NodeCount", "log10DBH"),
              ordtraits = c("log10DBH", "bio12", "bio15", "log10NodeCount"),
              dir_create = "results_BayesTraits_phyreg_input",
              fileDistData = "BayesTraits_linked_data_bio12_bio15_nnodes.txt",
              fileMeanData = "BayesTraits_mean_data_bio12_bio15_nnodes.txt",
              fileOrigData = "vataireoids_1610_25May2022_BayesTraits_netnodes_logtransf.csv")

2. phyreg.shell: Creating shell script for running phylogenetic regression analyses

Using directory paths to the tree and the input files of mean and linked trait data, this function produces shell scripts in both Unix and Windows PowerShell formats for running one or multiple phylogenetic regression analyses. This function creates the necessary batch file commands, including the desired settings for the burnin, iterations, and stepping stones parameters (e.g., abbreviated as bi, it, sa, and st) for a given run, in addition to incorporating any model of tree transformation. In the example below, we use “UNI”, “Lambda”, and “VR” in the treefransf argument to indicate we want three separate sets of batch file commands all identical except one, designated as “UNI” will not estimate “Lambda” or “VR” (variable rates). We use “UNI” to indicate the assumption of uniform rates of evolution and no other tree transformation (e.g., such as estimating “Lambda”). In this case, “UNI” here will consider a default BayesTraits regression analysis where “Lambda” assumes the default setting of 1.0 (see the BayesTraits manual).

Example of using phyreg.shell: 
library(InNOutBT)

phyreg.shell(meanfile_dir = "results_BayesTraits_phyreg_input/03Jun2022",
             linkfile_dir = "results_BayesTraits_phyreg_input/03Jun2022",
             treefile_dir = "Data",
             BayesTraits_dir = "/Users/domingoscardoso/BayesTraitsV4",
             responvar = "DBH",
             treetransf = c("UNI", "Kappa", "Lambda", "VR"),
             bi = "30000000",
             it = "130000000",
             sa = "100000",
             st = c("500", "50000"),
             syst = "unix",
             dir_create = "results_BayesTraits_phyreg_shell",
             cc_DataTree = TRUE)

3. phyreg.outputs: Processing output data from phylogenetic regression analyses

This function processes the output Log and Stones .txt files from phylogenetic regression analyses. The resulting processed files are tables either in Word or CSV formats summarizing each model, including comparisons of Bayes Factors. Heatmaps using pairwise Bayes Factor comparisons are also produced in PDF format (we use heatmaps only to identify models similar to the model with the highest likelihood score).

Example of using phyreg.outputs: 
library(InNOutBT)

phyreg.outputs(logst_dir = paste0(getwd(),"/BayesTraits_phyreg_outputs_log_stone"),
               responvar = "DBH",
               explanvar = c("bio12", "bio15", "nnodes"),
               treetransf = c("UNI", "Kappa", "Lambda", "VR"),
               bayesfactor = TRUE,
               unirates = FALSE,
               outformat = c("Word", "CSV"),
               tableleg = "Phylogenetic regression in BayesTraits: models and coefficients.",
               dir_create = "results_BayesTraits_phyreg_output",
               outfile = "BayesTraits_phyreg_output_table",
               height = 17,
               width = 20)

Documentation

A detailed description of the InNOutBT’s full functionality is available in different articles.

Citation

Cardoso, D. & Lavin, M. (2022). InNOutBT: An R package for creating input data and processing output data required for BayesTraits analyses. https://github.com/dboslab/InNOutBT