Gene blast identity problem for gapfilling

[feiranl]

Probelm in Gap-filling:

model yeast8
rxnA  Y...A

draft model:
rxnA gene1
gene1 Y...A blast Pidentity 45%
gene1 Y...B blast Pidentity 70%

however, Y...B doesn't link to any rxn in yeast8. Should we add this rxnA from draft model back to the specific models?

Another case:

model yeast8
rxnA  Y...A

draft model:
rxnA gene1
gene1 Y...A blast Pidentity 40%
gene1 Y...B blast Pidentity 30%

If the bi-directional pidentity is smaller than 45%, can we add the missing rxn back?