Interconversion between two unrelated protein folds in the lymphotactin native state

This could be a problem for de novo protein structure prediction?

---

What I'm trying to say is that if a protein has two different native-state structures (like the Lymphotactin) this could be a problem for the structure prediction.
Or in this case the Rosetta software will predict the two different structures correctly?

---

Rosetta, at present, would arrive at a prediction close to one of the two possible native structures. But that's ok for a couple of reasons. If the protein is a virus you wish to create a vaccine for, if your vaccine presents itself with an outer surface that is similar on both of the possible native structures. If there is not enough surface area between the two de novo structures, you could use Rosetta to design two vaccines. One that will present itself as each of the possible forms.

Similarly for docking. If you are attempting to bind directly to a viral protein, you could design two unique binding proteins and use Rosetta to model how each will bind to it's designed target. Or, you could design a protein that will bind to the area of the protein that are common between the two.

---

Rosetta Moderator: Mod.Sense

Thanks for the answer.
I understand that in case of real problems the issue can be resolved in other way.
But I'm thinking at the theoretical model and I have an other question.
If I have understood correctly, the Rosetta software search the lowest point in the energy landscape, but in this case there would be two lowest point.
Rosetta would find both of them? It would be interesting to make a structure prediction for the Lymphotactin and see the results.

Just my two cents. :)

---