The training set had shapes that the proteins usually take - which is a result of interactions with the things that usually are in their environment (including

The training set had shapes that the proteins usually take - which is a result of interactions with the things that usually are in their environment (including copies of itself). So to correctly predict that - the algorithm had to learn to simulate these interactions, right?