has participant property comes in the wrong way

[matentzn]

Some downstream tools get a bit confused by this:

drosophila-phenotype-ontology/dpo-full.obo

Line 15821 in a746a8d

[Typedef]

(There already is a correctly mapped has participant role). The root of the problem is CARO, which contains http://www.obofoundry.org/ro/ro.owl#has_participant in its signature. Should be fixed at some point.