“…There is, nevertheless, a preferred axis at the location of each main chain sphere corresponding to the tangent along the chain or the direction along which the chain is oriented at that location. Replacing the spheres with objects such as unidirectional coins and uniaxial discs, allowing neighboring spheres to overlap, or adding side chains to the spheres along the main chain are all steps that break the spherical symmetry and yield ground state structures, which resemble protein structures to varying degrees [ 51 , 52 , 53 , 54 , 55 , 56 , 57 , 58 , 59 , 60 , 61 ]. To avoid clouding the issue by studying an approximate model for proteins, we resort instead to a careful analysis of experimental data of over 4000 protein native state structures (see Section 2 ).…”