Ramachandran plots of amino acid residues of short peptides show that the sampled conformational space is smaller than assumed by the classical random coil model. At variance withthe latter, it depends on the type of nearest neighbor in a polypeptide chain. . Considering nearest neighbor interactions is pivotal for reproducing conformational distributions of unfolded and intrinsically disordered proteins.

Intrinsically disordered and artificially denatured foldable proteins have many properties in common, even though the distribution of amino acid residues generally differ. Both are traditionally described as either adopting molten globule or random coil-like structure, depending on their amino acid composition. For a long period of time, it was believed that with the exception of glycine and proline, conformational preferences of individual amino acid residues are very similar, and therefore, not of great concern. This article discusses experimental and bioinformatical data which show that individual Ramachandran plot distributions in unfolded proteins depend on the characteristics of side chains. Furthermore, ample experimental evidence suggests the presence of nearest neighbor interactions between residues which are ignored in the classical random coil model. The article reviews and discusses work that emphasizes the necessity of a thorough understanding of the local behavior of disordered and unfolded proteins alike for an understanding of the coiled state of proteins. Unresolved issues are delineated and research perspectives suggested.