2  Requirements

2.1 Install R

If on Windows or macOS, go here to install R.

If on Ubuntu or another Debian-based Linux installation, install r-base and r-base-dev so that you can compile R packages:

sudo apt-get install r-base r-base-dev

2.2 Linux system packages

Several R packages will need to be installed. If you are using Linux, the R packages will likely be installed from source. Therefore, you need to install a few Linux system packages that are required:

  • GSL (GNU Scientific Library). Instructions for installing GSL are here. On most Linux distributions, it should be installable with sudo apt-get install libgsl-dev.

  • GraphicsMagick. On most Linux distributions, it should be installable with sudo apt-get install graphicsmagick.

  • Many system packages are needed for the devtools R package. These can be installed with sudo apt-get install gfortran build-essential libcurl4-openssl-dev libxml2-dev libfontconfig1-dev libharfbuzz-dev libfribidi-dev libfreetype6-dev libpng-dev libtiff5-dev libjpeg-dev.

2.3 R packages from CRAN

The required R packages hosted on CRAN (Comprehensive R Archive Network) can be installed by initiating an R session and inputting

install.packages(c("devtools", "data.table", "RCurl", "RJSONIO", "future",
  "future.apply", "parallelly", "lubridate", "collapse", "actuar",
  "gbutils", "binsmooth", "spatstat.univar", "distributionsrd",
  "VGAM", "VaRES", "ghyp", "extraDistr", "GB2", "wrswoR", "Rfast",
  "fitdistrplus", "bsgof", "PearsonDS", "animation", "triangle",
  "viridis", "RColorBrewer", "ISOweek", "ggh4x"), Ncpus = 4)
Important

R may ask you if you want to create a personal library directory to install packages into. Select “yes”.

2.4 R packages from GitHub

One package must be installed from GitHub:

remotes::install_github("dracula/dRacula", upgrade = FALSE)

2.5 OSPEAD R package

In an R session launched from a directory that contains decoyanalysis_0.1.0.tar.gz, input:

install.packages("decoyanalysis_0.1.0.tar.gz", repos = NULL)

2.6 Fast BLAS

Enabling a fast BLAS (Basic Linear Algebra Subprograms) is strongly recommended. Instructions for Ubuntu and Windows are here. On Ubuntu, sudo apt-get install libopenblas-base should install and enable the fast BLAS.

With Linux, to check that the fast BLAS has been enabled, open a new R session and input sessionInfo(). One of the messages should be

Matrix products: default
BLAS:   /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3
LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/liblapack.so.3

2.7 Multi-machine

It is possible to run the most computationally-demanding parts of the code on multiple machines in parallel. These OSPEAD-docs instructions assume a single machine. Multi-machine instructions may be added later.