Map/axiomtize NCBI metagenomes hierarchy with ENVO

[cmungall]

E.g.

/ NCBITaxon:1 ! root
  is_a NCBITaxon:12908 ! unclassified sequences
   is_a NCBITaxon:408169 ! metagenomes ***
    is_a NCBITaxon:1235509 ! synthetic metagenome
    is_a NCBITaxon:256318 ! metagenome
    is_a NCBITaxon:410656 ! organismal metagenomes
     is_a NCBITaxon:1009879 ! Crustacea gut metagenome
     is_a NCBITaxon:1041057 ! sea squirt metagenome
     is_a NCBITaxon:1070528 ! viral metagenome
     is_a NCBITaxon:1076140 ! termite metagenome
     is_a NCBITaxon:1077527 ! beetle metagenome
     is_a NCBITaxon:1077528 ! ant metagenome
     is_a NCBITaxon:1077529 ! wasp metagenome
     is_a NCBITaxon:1115523 ! upper respiratory tract metagenome <-- UBERON/ENVO
...
     is_a NCBITaxon:540485 ! mouse skin metagenome
     is_a NCBITaxon:646099 ! human metagenome
     is_a NCBITaxon:662107 ! phyllosphere metagenome       
     is_a NCBITaxon:703336 ! wallaby gut metagenome <-- UBERON+TAXON
     is_a NCBITaxon:749906 ! gut metagenome
     is_a NCBITaxon:910258 ! honeybee metagenome
    is_a NCBITaxon:410657 ! ecological metagenomes  <-- ENVO 
     is_a NCBITaxon:1006967 ! shoot metagenome
     is_a NCBITaxon:1070537 ! biofilter metagenome
     is_a NCBITaxon:1076179 ! bioreactor metagenome
     is_a NCBITaxon:1082480 ! permafrost metagenome
     is_a NCBITaxon:1118232 ! root metagenome
     is_a NCBITaxon:1154581 ! food fermentation metagenome 
     is_a NCBITaxon:1169740 ! aquatic metagenome
...

[cmungall]

Consider also axiomatizing the enumerations in the Core schema for bioprojects. For more info see:
https://www.biostars.org/p/44805/

[pbuttigieg]

That is interesting, and also in need of semantics. For example, contrast 'viral metagenome' with 'beetle metagenome' (and most others) - the former references the target of investigation while the latter the 'host' or environment of the metagenome.

In relation to the NCBITax import - is there any way to make it less memory hungry?
Looking at this