Public data source

I need to create a public Zenodo repository to work in the same way as the current compendium dataset—downloadable files used by both the JS compiler and R package. Files needed (after implementation of https://github.com/blekhmanlab/compendium_website/issues/39)

* loadings.txt – taxon weights for all ordinations
* sample-pcs.tsv – projected coordinates for all compendium samples
* scree.tsv – variance explained by the first axes of each projection
* *Example data* (for each project):
* example1.asv.txt
* example1.taxa.txt
* example1.meta.txt
* example1.projected.txt (for validation)

Plus for project similarity search:
* project-pcs.tsv – project centroids
* projects.txt – project-level metadata