Do we maintain the subclass of (P279) property in the gene model?

I have updated the ShEx model for the human genes. The initial model contained both a P31 and P279 property. Initially the p31 property had only the gene value (i.e. Q7187) and the p279 captured the different types of genes. Currently, these types of genes are directly captured with the P31 property.

Initially, we had:

p:P31 @<P31_instance_of_gene> ; # Item describes a gene
  p:P279 @<P279_subclass_of_gene>+ ; # Item is a subclass of e.g protein-coding gene.
<P31_instance_of_gene> { 
   ps:P31 [wd:Q7187] ;     # Instance of [P31] gene [wd:Q7187]
   prov:wasDerivedFrom @<ncbi-gene-reference> OR @<ensembl-gene-reference> ;
} 
<P279_subclass_of_gene> { 
    ps:P279 @<gene_types> ; # Subclass of [P279] gene types <gene_types>
    prov:wasDerivedFrom @<ncbi-gene-reference> OR @<ensembl-gene-reference> ;
} 
<gene_types> [
  wd:Q7187  # gene
  wd:Q20747295 # protein-coding gene
    ]

This changed into

p:P31 @gw:P31_instance_of_gene ; # Item describes a gene
gw:P31_instance_of_gene { 
   ps:P31 @gw:gene_types ;     # Instance of [P31] gene types
   prov:wasDerivedFrom @gw:ncbi-gene-reference OR @gw:ensembl-gene-reference ;
} 
gw:gene_types [
  wd:Q7187  # gene
  wd:Q20747295 # protein-coding gene
  wd:Q427087 # ncRNA
  wd:Q284578 # snRNA
  wd:Q284416 # snoRNA
  wd:Q215980 # rRNA
  wd:Q201448 # tRNA
  wd:Q277338 # pseudo
  wd:Q11053 # miscRNA
  wd:Q25323710 # scRNA
             ]

In the latest gene model, the subclass of no longer exists. The question is if this is accurate? If not, how should the subclass of shape look like? Also if it is accurate, does it make sense to capture subclass as an unacceptable shape?

SuLab / GeneWikiCentral

Do we maintain the subclass of (P279) property in the gene model? #81