Food allergies tend to be a response to one compound, or a very small set of compounds. With respect to using genes from one organism to confer hardiness on another, the chances of conferring the production of a deadly allergen are diminishingly slim, but you'd better believe that if such a thing was to be done, the FDA (or analogous organizations outside the US) would have warnings plastered all over the derivative organism.

The level of justification and background research showing how you're NOT going to destroy the world that is required to even get funding for this sort of thing is.... large.