Trainers: Iris Yu, Jiawei Wang, Anna Vathrakokili, Yuyao Song, Andrian Yang, Nadav Yayon
Contributor: Hugo Tavares
To run the scripts for generating counts from raw fastq files Demonstrations/01_*.sh, please install CellRanger using these instructions.
To run all the other demo notebooks, follow the instructions below. It walks you through creating and using a container that has all the packages needed to run the demo notebooks.
- Install Docker or Singularity
- Singularity is best for an HPC environment as a regular HPC user. Check if your institute's HPC already has it.
- Docker might be easier to install in your local machine, depending on your machine.
- Pull the prebuilt training image that Andrian Yang built, here. This image is based on the official scanpy image gcfntnu:/scanpy:latest.
- If using Signularity, do:
wheresingularity pull scrnaseq2024.sif docker://ghcr.io/andr-kun/scrnaseq2024-container:latestscrnaseq2024.sifis the name you would like to save the image file as.- If using Docker, do:
docker pull ghcr.io/andr-kun/scrnaseq2024-container:latest- Alternatively, build the image using Andrian’s Dockerfile.
- Run Jupyter lab using the image you built, either using Docker or Singularity. If Singularity, you run it the same way we call it during the course sessions.
- Using Singularity, run the command below while standing in this repo's root directory, or whichever directory you would like to be the root of Jupyter when it initialises.
singularity exec scrnaseq2024.sif jupyter lab- Using Docker, run the command below while standing in this repo's root directory. The argument
-v "$(pwd):$(pwd)"will bind mount the present working directory to the container such that you can call it as is, and-w "$(pwd)"will set the present working directory as the working directory in the container. Jupyter will initialise with the present working diretory as the Jupyter root.
docker run -p 8888:8888 -v "$(pwd):$(pwd)" -w "$(pwd)" ghcr.io/andr-kun/scrnaseq2024-container:latest jupyter lab --ip 0.0.0.0 --allow-root