Martinotti cells no longer classify

[memartone]

I suspect it is because the -PV and -VIP have been added as necessary and sufficient conditions: Mammalia neocortex (unionOf L2 L3 L5) Martinotti +Sst -PV -VIP +GABA intrinsic neuron. Those will have to be removed again, although I don't know why Markram doesn't classify under the Martinotti CUT.

[tgbugs]

The negative phenotypes need to be moved to entailed because they are
too strong to ever be satisfied by experimental evidence (they are
open). On the other hand, they are excellent webs for catching pesky
types that to claim the have both a certain molecular phenotype and
morphological phenotype, where that pairing has been deemed to be
impossible.

Essentially strong negative phenotypes that combine a property that
has many subproperties with an object that has subclasses that are not
closed, will only satisfy when the exact pairing or an even strong
pairing is used, e.g. hasPhenotype with a molecular indicator.

Molecular indicators should almost never be used directly in EBMs,
though in some cases one might be able to say "we looked everywhere
and could find no PV except in some compacted chormatin DNA which we
take as a negative indicator, and also FISH and antibody didn't show
up, so we are going to make a very strong statement."

Thus the closest we might get is an equivalence between experimental
observations that we claim imply the negative phenotype directly.
The gene being in compacted chromatin and the molecule not being a
neurotransmitter might be assert to be equivalent to the negative
phenotype, allowing room for scientific and evidence based review of
whether that seemed to bear out in experiments.