Many of the datafiles are coming from publications, while others are generated by scripts. Ideally these should be documented in this repo.