Add torben conversion script along with the output data

[jayrbolton]

No description provided.

[jayrbolton]

@realmarcin I just updated this PR to separate data into its own directory, and added some minimal documentation.

[realmarcin]

can you align this file naming to:
IMGVR_extra_KGX_edges.tsv
IMGVR_extra_KGX_nodes.tsv

so subKG (Torben) dataset (sample) _KGX_edges.tsv etc

Also the format is KGX (and overlaps with many graph formats) and we are using some biolink terms to annotate our nodes/edges.

[realmarcin]

A few notes for the edges file:

• please leave out lat/long for now since we can't ingest values directly (we have pairwise sample distance data, we could try to bin that but separate ingest ok there)
• All the IMG: prefixes should just be GOLD: (for now, we can deal with the few IMG ids later)
• some stray asterisk here GOLD:*Microbiome?
• Your primary (subject) id should start with Ga (for GOLD analysis sample id) so maybe currently its not the right column eg IMG:403934
• 'located_in' for now is just strictly geography, so all other environment labels just get has:attribute
• Assembly_size and Gene_count values need to be log10 to get the bins
• Not sure where these are coming from, but possibly can remove -- unless its the same field as in IMGVR_to_tsv.py:
  IMG:403936 biolink:has_attribute GOLD:DOE Joint Genome Institute (JGI) biolink:has_attribute
• I think here you are using TaxonOID? But that's not an actual taxon id, so these also need to be removed:
  IMG:403936 biolink:has_taxonomic_rank IMG:3300021138 biolink:has_taxonomic_rank

[realmarcin]

A few notes for the nodes file:

• The edges comment apply here too, most of the issues I think.

• This one looks odd, no prefix? Not sure what this field is but also mismatch between id and label:
  WI - Practice 03JUN2009 epilimnion biolink:NamedThing

• Looks like all nodes are type biolink:NamedThing? If you follow the IMGVR sample script there are more specific categories for all but two cases I think, take a look.

• Also note the field regex to clean them up, eg remove white spaces, applies to both nodes and edges -- I just do it on the entire data frame:
  df = df.replace(',', '', regex=True)
  df = df.replace(' ', '', regex=True)
  df = df.replace(':', '', regex=True)
  df = df.replace('__', '', regex=True)

[realmarcin]

here let's align to the KGX dir structure:
ls -1 data/
raw
transform
merged

and then subdirs for each source as in the merge.yaml
https://github.com/kbaseIncubator/KE_KG/blob/main/merge.yaml