is hepatic duct intrahepatic part always an arborising epithelium?

[cmungall]

Superclasses:

   is_a CARO:0000043 ! simple tissue
     is_a CARO:0000066 ! epithelium
      is_a AEO:0000115 ! epithelial sac
       is_a AEO:0000117 ! arborising epithelial duct system
        is_a EHDAA2:0000743 ! hepatic duct intrahepatic part ***  [ends_at: "CS20"] [starts_at: "CS13"]

Is the is_a correct here?

Gives rise to:

                  df EHDAA2:0000941 ! left hepatic duct [ends_at: "CS20"] [starts_at: "CS14"]
                   df EHDAA2:0001002 ! liver left lobe parenchyma [ends_at: "CS20"] [starts_at: "CS14"]
                  df EHDAA2:0001727 ! right hepatic duct [ends_at: "CS20"] [starts_at: "CS14"]
                   df EHDAA2:0001007 ! liver right caudate lobe parenchyma [ends_at: "CS20"] [starts_at: "CS17"]
                   df EHDAA2:0001010 ! liver right lobe parenchyma [ends_at: "CS20"] [starts_at: "CS14"]
                   df EHDAA2:0001014 ! liver right quadrate lobe parenchyma [ends_at: "CS20"] [starts_at: "CS18"]
~

Is EHDAA2 intended for reasoning about non-epithelial layers of ducts? These would have different origins presumably. I think for Uberon we should read many EHDAA2 duct terms as being 'duct epithelium'. It might be an idea if EHDAA2 made this explicit in the label.

Original ticket:
obophenotype/uberon#654