Machine Learning for Biomedical Science: Setup

Project organization

1. Start RStudio.

2. Create a new project in your Desktop called ml-biomed.
   • Click the File menu button, then New Project.
   • Click New Directory.
   • Click New Project.
   • Type ml-biomed as the directory name. Browse to your Desktop to create the project there.
   • Click the Create Project button.

3. Use the Files tab to create a data folder to hold the data, a scripts folder to house your scripts, and a results folder to hold results. Alternatively, you can use the R console to run the following commands for step 2 only. You still need to create a project with step 1.

   dir.create("./data")
   dir.create("./scripts")
   dir.create("./results")
   

Package installation and data access

1. Install R packages and load the libraries. You might already have some of these installed. Search for their names in the Packages tab in RStudio. There is no need to install them again if you find them there.

   install.packages("devtools", "rafalib", "RColorBrewer", "gplots", "UsingR", "class", "caret")
   library(devtools)
   library(rafalib)
   library(RColorBrewer)
   library(gplots)
   library(UsingR)
   library(class)
   library(caret)
   

2. Install the tissue gene expression data from Github then load the library and data. If you were successful installing and loading devtools above, you should be able to install from Github.

   install_github("genomicsclass/tissuesGeneExpression") 
   library(tissuesGeneExpression) 
   data(tissuesGeneExpression)
   

   The data represent RNA expression levels for eight tissues, each with several individuals.

   Alternatively, download the data directly from Github and place them in your new data directory. Do this if installing devtools or installing from Github gave you difficulty.

3. Load packages from Bioconductor.

   if (!require("BiocManager", quietly = TRUE))
   install.packages("BiocManager")
      
   BiocManager::install(c("genefilter", "Biobase", "SpikeIn", "hgu95acdf")
   library(genefilter)
   library(Biobase)
   library(SpikeIn)
   library(hgu95acdf)
   data(SpikeIn95)