Figure 1. Importance of IDPs in different organisms. Structural disorder has been predicted in different groups of organisms by the IUPred algorithm (see Table 3). The percent of residues predicted to be disordered in proteins has been averaged for species and then for the whole taxonomic group. The results show a significant difference between prokaryotes and eukaryotes, with a high level of disorder in eukaryotes (Tompa et al., unpublished observations).

Figure 2. Schematic representation of different kinds of IDP functions. The diagram shows the two basic types of IDP functions (Table 1).

1. An IDP may exert its effect directly due to its disorder, when it acts as an entropic chain. There are different types of such function; here it is shown that several disordered chains binding to the central protein exert long‐range entropic repulsion (entropic bristle) preventing other macromolecules from coming close.
   
2. The other basic type of IDP function is binding to a partner, when the IDP undergoes induced folding and assumes a structured state with various functional consequences.
   

Figure 3. Dots in the graphs represent the expected correlations involving backbone nitrogens with the directly bound amide proton (N ^H _i–H ^N _i) or carbonyl carbon (N ^H _i–C′ _{i – 1} ), calculated using random‐coil chemical shifts appropriately corrected for the contributions from the primary sequence, taking human securin as an example. For a fair comparison, the same spectral widths (in units of hertz) are shown in the two graphs (2 ppm for corresponds to 8 ppm for ). The HN HSQC and CON experiments acquired on intrinsically disordered human securin are reported in the original publication [ 72].

Figure 4. Comparison of – HSQC (top) and – CON (in‐phase/antiphase) (bottom) spectra. The – HSQC and – CON‐IPAP spectra of human securin are shown to illustrate the additional information content from protonless NMR experiments. In the – CON‐IPAP spectrum, the circle indicates the Pro residues. The first count of cross‐peaks gives 122 in HSQC (68% of the expected, 60% of the whole protein) and 165 in CON (82% of the expected, 82% of the whole protein).

Figure 5. Chemical shifts as indicators of secondary structural elements. As an example, the difference of experimental chemical shifts (C ^α–C ^β in this case) respect those predicted using random coil chemical shifts, properly corrected for the contribution by the primary sequence Δ(C ^α–C ^β), values are reported for different proteins with similar size, but with different structural and dynamic properties. From left to right: Atx1 (73 amino acids) characterized by a well‐defined three‐dimensional structure, and Cox17y (69 amino acids) characterized by heterogeneous redox‐state‐dependent structural and dynamic behavior with a largely folded oxidized state in which the SS bonds stabilize the helix–coil–helix fold and a largely disordered reduced state.

Figure 6. – steady‐state NOE values as a quick indicator of regions characterized by markedly different motional properties. (a) Atx1 (73 amino acids) characterized by a well‐defined three‐dimensional structure. (b) Cox17y (69 amino acids) characterized by a heterogeneous redox‐state‐dependent structural and dynamic behavior with a largely folded oxidized state in which the SS bonds stabilize the helix–coil–helix fold and a largely disordered reduced state. For clarity a plot of the expected – steady state NOE values as a function of the correlation time (calculated at 500 MHz) is also reported.