Clone this repo to your computer.
git clone git@github.com:UCSF-DSCOLAB/hackathon_primer.git
OR
git clone https://github.com/UCSF-DSCOLAB/hackathon_primer.git
This repo contains the Dockerfile and sample data with tutorial used to create the hackathon containers for single-cell analysis for both python and R-studio.
The container comes pre-installed with basic single cell tools in Python and R. Sample Data with tutorial also included.
Users can start analyzing single cell sequencing data with Scanpy in python or Seurat in R.
- Docker
Note: The easiest way to install and use docker, is via docker desktop: https://www.docker.com/products/docker-desktop/
Be sure to select the appropriate installation for you Macbook machine.
After installing docker, pull the repository from docker hub.
docker pull drbueno/single-cell-nb:latest
Change directory to python-container.
Run (recommended to run under screen)
./start.sh
You will be prompted to set your working directory. This is the directory where the data lives.
ALL work must be done in this directory. It will be mounted inside the container in /home/data
An example of working directory: /Users/hackathon-user/UCSF_HACKATHON_PRIMER/python-container/data
After entering the path of your local work directory, follow instructions to copy and paste link with IP address to a web browser.
Be sure to have:
- The latest version of Docker
And in Docker Settings (using docker desktop):
- General -> User Virtualization Framework -> ON
- Features in development -> User Rosetta for x86/amd64 emulation on Apple Silicon -> ON
The Jupyter Notebook with Scanpy tutorial for Preprocessing and clustering 3k PBMCs path is here:
/home/tutorial/
Open Jupyter notebook file python-tutorial.ipynb and run tutorial for python.
Install conda. https://docs.conda.io/projects/conda/en/latest/user-guide/install/index.html
Change directory to python-container.
Create single-cell conda environment.
conda env create -f env.yaml -n single-cell
Activate single-cell environment.
conda activate single-cell
Start a jupyter notebook (recommended to run under screen).
jupyter lab
The jupyter notebook will have both python and R kernel. Choose what kernel you want to work with and start analyzing your data.
After installing docker, pull the repository from docker hub.
docker pull drbueno/rstudio-single-cell:latest
Change directory to rstudio-container.
Run ./start.sh
You will be prompted to set your working directory. This is the directory where the data lives.
ALL work must be done in this directory. It will be mounted inside the container in /home
An example of working directory: /Users/hackathon-user/UCSF_HACKATHON_PRIMER/rstudio-container/data
Go to a web browser and visit
localhost:8787
Log in with username rstudio and password @hackathon2023
To see the tutorial, open the .Rmd file in /tutorial
Clone the rstudio-container repo.
Change directory to repo, open the command line and run
Rscript install_monocle_and_R_dependencies.R
Your local R or R-studio should have all packages installed needed for single-cell analysis.
If you are going to install other packages, DO NOT stop your containers. If you stop your
containers, the packages that were installed will not exist when you restart the container.
Instead run the ./start.sh
under screen. When the Hackathon is complete, stop the container.
When the Hackathon is complete. You can stop the container from running.
To see all containers running, type in command line docker ps
To stop containers run docker stop <CONTAINER ID>
- Click docker app icon on your computer.
- Click on the setting icon
- Once in settings, click on Resources
- Change the parameters based on need (in the 2021 Hackathon, memory was an issue.)
