Energy Landscape of a Model Protein

The potential energy surface of an off-lattice model protein is characterized in detail by constructing a disconnectivity graph and by examining the organisation of pathways on the surface. The results clearly reveal the frustration exhibited by this system and explain why it does not fold efficiently to the global potential energy minimum. In contrast, when the frustration is removed by constructing a `Go-type’ model, the resulting graph exhibits the characteristics expected for a folding funnel.