INSTALL.md

In farrellja/URD: URD

Installing URD:

Quick Installation

1. Install R and RStudio

URD is a package written in R and designed to be used in the RStudio interactive environment.

R can be obtained and installed from https://cran.rstudio.com.

Following installation of R, Rstudio can be obtained and installed from https://www.rstudio.com/products/rstudio/download/. The free version of RStudio Desktop is sufficient.

2. Install an X11 window client (Mac only)

3D plots produced by URD require an X11 windowing client to display. Windows can do this natively, but Mac requires additional software. Our favorite for Mac is XQuartz (https://www.xquartz.org). After installation, restart your computer.

3. Install required R packages and URD

We wrote a script that will attempt to install all requirements for URD and then URD itself. It has been tested on R 3.5 and R 3.6. Start RStudio and then run the installation script from the console:

source("https://raw.githubusercontent.com/farrellja/URD/master/URD-Install.R")

Troubleshooting

A separate conda environment

If you have trouble installing URD because of version conflicts with your currently installed packages, you can create a separate conda environment for URD:

conda create urd base-r==3.5.1 udunits2 -c conda-forge 
conda activate urd
Rscript -e 'source("https://raw.githubusercontent.com/farrellja/URD/master/URD-Install.R")'

udunits

URD's dependencies require the installation of the udunits2 package, which depends on the udunits2 library. This is part of the Windows and Mac OS X operating systems, but often requires manual installation on Linux distributions:

sudo apt-get install libudunits2-dev

Or if using conda:

conda install udunits2

DLL error

R has limit on the number of DLLs that can be loaded by linked packages. If you receive a maximal number of DLLs reached error during installation, then you can increase the number of simultaneously allowed DLLs from 100 to 200 by modifying the .Renviron file.

On Mac, from the terminal, run: echo "R_MAX_NUM_DLLS=200" >> ~/.Renviron

Manual installation

If something went wrong previously, you may have try installing some of URD's dependencies manually:

A. Install and attach the devtools package

devtools is required to compile and install URD, since it's distributed through GitHub.

install.packages("devtools")

B. Install required Bioconductor packages

Because these packages are deposited in Bioconductor and not CRAN, they must be installed manually before installing URD (otherwise its installation will fail.)

source("https://bioconductor.org/biocLite.R") biocLite(c('Biobase', 'S4Vectors', 'AnnotationDbi', 'destiny'))

Optionally, additional packages that are required for specific functions can be installed.

biocLite(c('sva', 'rhdf5', 'scran'))

If installing on a cluster, where you do not have write permissions to the main R libraries (for instance, if you receive errors like "ERROR: no permission to install to directory", you may need to specify a library location, such as:

biocLite(c('sva', 'rhdf5', 'scran'), lib=[PATH TO DIRECTORY TO STORE R LIBRARIES WHERE YOU HAVE PERMISSION TO WRITE])

C. Install URD

URD can be installed directly from the GitHub repository

library(devtools) install_github(repo="farrellja/URD")

farrellja/URD documentation built on June 17, 2020, 4:48 a.m.