You can select from our current recommended image pranavmishra90/facsimilab-full:latest. It contains a conda (micromamba) based python environment which follows our fully functional bare metal (non-Docker) environment.
-
facsimilab-base: Adds functionality for python (micromamba) and a number of apt packages into the CUDA capable
nvidia/cuda:12.1.0-base-ubuntu22.04 -
facsimilab-main: Creates a python 3.11 base environment with the essentials for statistics, graphing, documention, and reproducible science with
datalad,quarto, andrclone -
facsimilab-full: Creates the python 3.11 full facsimilab environment with a large number of packages capable of completing a variety of experiments, including:
- Clininal research with REDCap:
pycap - Next generation -omics:
scvi,scanpy,gseapy,pydeseq2,celltypist, etc. - Machine learning:
scikit-learn,leidenalg,imbalanced-learn - Reproducible research (file versioning, archival, and documentation):
datalad,git,git-annex,rclone,quarto jupyternotebooks withpapermillautomation
- Clininal research with REDCap:
- Micromamba: A lightweight form of mamba, which itself is far faster than conda in creating python virtual environments
- Datalad - version controlling large datasets
- Git-Annex - included with Datalad
- Quarto - generate documentation programmatically
- Rclone - add additional git remotes (siblings) for
datalad - Nvidia CUDA - GPU accelerated analysis
You can quickly deploy FacsimiLab using the docker run commands found in Quick Deploy. For futher testing, a docker-compose.yaml file is available in /testing/.
cd docker
bash build-all.shCopyright (c) 2022-2024 Pranav Kumar Mishra
Licenses and references of open-source projects that contribute significantly to FacsimiLab are listed in Licenses