Using a Personalized, Adaptive and Cooperative MultiAgent System to Predict Protein Secondary Structure