After discussions resulting from experiments with shex validation, the team has decided to add the assertion rdf:type Molecular Function to the currently untyped nodes representing transport processes.
Note that there is a proposal to add a new class called something like Reaction as a parent of Molecular Function in the GO and use this class instead. The gist being that these processes shouldn't be called Molecular Functions because they have no enabling gene but, at the same time, are biologically meaningful and might best be described as reactions. So the story may continue.
After discussions resulting from experiments with shex validation, the team has decided to add the assertion rdf:type Molecular Function to the currently untyped nodes representing transport processes.
See history on these challenging creatures in this thread https://github.com/geneontology/pathways2GO/issues/17
Note that there is a proposal to add a new class called something like Reaction as a parent of Molecular Function in the GO and use this class instead. The gist being that these processes shouldn't be called Molecular Functions because they have no enabling gene but, at the same time, are biologically meaningful and might best be described as reactions. So the story may continue.