Structuring a data science project

It is important to have a clear structure of the project, such that several people can work together

• README.md (description of project)

• .gitignore (we do not want to put data/pdbs and data/fasta under git control)

• notes (documentation and notes, just use reasonable filenames)

• data

  • ab_ag.tsv, ...
  • pdbs (downloaded pdbs data)
  • fasta (created fasta files)

• notebooks

  • preprocess (QC, data cleaning)
  • contact_points
  • SASA
  • analysis1
  • analysis2

• generated (data that will be kept)

  • preprocess (outputs from notebooks/preprocess)
  • contact_points (compute contact points)
  • SASA (compute SASA)
  • analysis1 (first analysis)
  • analysis2 (second analysis)

The idea of having the generated folder is that we can modularize the analysis and development.