-
Notifications
You must be signed in to change notification settings - Fork 0
Galen new user setup
After our move to the University of Pittsburgh, we have been granted a grace period on galen until approximately September 2024, after which we will no longer have access to this resource.
Follow the Unix 101 tutorials here:
- Part 1: https://drive.google.com/file/d/1L4Fzb327c0HSn4RmIqdyh6wwx0AhbOYg/view
- Part 2: https://drive.google.com/file/d/1GyCnjHHfGKthCcYK6cWxviMAD3suZDi6/view
- Part 3: https://drive.google.com/file/d/1SgamG47eg1nKtHMFo8IUPujfQsIbX5Bp/view
- Part 4: https://drive.google.com/file/d/1xn9Wx9nbvWumgvn4O9d73XeCVLSW1sE5/view
- Part 5: https://drive.google.com/file/d/1O43OGkgT7WMNGo8MWsNxPOq1SBbGWN7R/view
Disable the gnu12 module and save this module state. Otherwise this will interfere with the default R installation.
module reset
module unload gnu12
module save
Modify your .bashrc to automatically restore this saved state every time you log-in. While we're in here, we can also add a python alias and a new path entry and also increase the number of files allowed to be opened simultaneously for bcl2fastq to work. Before we make any changes, let's make a backup copy just in case:
cp ~/.bashrc ~/.bashrc.backup
To edit the file, use nano (or if you know what you're doing, use vim or any editor of your choice).
nano ~/.bashrc
Then add the following lines to the end of the file:
case "$HOSTNAME" in
#load custom default module state (with gnu9 module disabled)
galen*) module restore;;
#increase number of file handles allowed for bcl2fastq
rothseq1*) ulimit -n 2048;;
esac
#create a python alias
alias python='python3.8'
#add the .local/bin folder to your PATH variable
export PATH=$HOME/.local/bin:$PATH
If you used nano, save with CTRL-S
and exit with CTRL-X
.
If .local/bin
(and bin/
) don't exist yet, make sure to create them now.
mkdir -p ~/.local/bin/ ~/bin/
#create a downloads folder
mkdir downloads
cd downloads
#clone the clusterutil repo
git clone https://github.com/jweile/clusterUtil.git
#change to repo folder
cd clusterUtil
#run installer
bash install.sh
#go back to the downloads folder
cd ~/downloads/
#download conda
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
#install conda
bash Miniconda3-latest-Linux-x86_64.sh
- Follow the on-screen prompts
- Now, log out (via
exit
) and in again (viassh
) to update the session
Even though it's technically already installed, the accompaying libraries are not well maintained by the sysadmin. So we install our own copy:
conda install python=3.8
#follow prompts again
#update all packages afterwards
conda update --all
Both at the LTRI and the Donnelly Centre a great number of people access Github on a daily basis all while sharing the same external IP address. From Github's perspective this can appear as a single IP address spamming it with access requests, thus throttling or temporarily banning access from our institutions. This is a problem when installing R packages directly from github, as this requires frequent access to it.
To get around those access limits, Github offers Personal Access Tokens (PATs), which looks like a string of random letters and numbers that serves as an application-specific password to your account. To learn how to create your PAT, refer to the relevant github help page.
Once you have generated a PAT, you can let R
know to use it for future installations.
- Go to your home directory and create a file called
.Renviron
cd ~
nano .Renviron
- Type
GITHUB_PAT=
and then copy-paste your PAT string after the=
sign. Finally, press Enter to add a new line at the end - Save and exit nano (
CTRL-S
andCTRL-X
)