Standard folder structure for data analysis projects in the BioMeDS group.
This is the basic high level folder structure. All these folders can and should contain sensible sub-folders. If you have different experiments or sub-tasks, try to use consistent sub-folder structures for data, code, results, etc.
.
├── code ← all script and notebook files
├── data
│ ├── processed ← any transformations of the raw data (scaling, normalization, ...)
│ └── raw ← raw data as received, read only
├── documents ← abstracts, papers, posters, grant proposals, talks, ...
│ └── references ← .bib file
├── envs ← compute environments (e.g. conda .yml or renv folder)
├── .git ← the project is obviously a git repository
├── .gitignore ← used to exclude temporary files and large files
├── LICENCE ← defaults: data: CC0, documents/figures: CC-BY, code: MIT
├── README.md ← high level documentation, links to analyses and summary of results
└── results
├── figures
├── final
└── intermediate
If you use tools for large data handling (datalad or dvc), containers (apptainer or docker), a workflow language (Snakemake), develop machine learning models or develop a software package (possibly with nbdev) as part of your project, the folder structure is extended accordingly
.
├── apptainer|docker ← container definition (.def or Dockerfile) and image (.sif)
├── code
├── data
│ ├── processed
│ └── raw
├── .datalad|.dvc ← handled by datalad or dvc
├── documents
│ ├── paper
│ ├── poster
│ └── references
├── envs
├── .git
├── .gitignore
├── LICENCE
├── <pkgname> ← subfolder for your package
│ ├── doc
│ ├── src
│ └── tests
├── README
├── results
│ ├── figures
│ ├── final
│ ├── intermediate
│ └── models ← exported model weights, code for loading the model in the `code` folder
└── workflow ← Snakefile(s) and config for Snakemake
└── Snakefile