FIG. 1.

(a) A schematic of a folded protein represented by a polymer of connected disks (${\mathrm{C}}_{\alpha }$ atoms) with non-bonded pair-wise ${\mathrm{C}}_{\alpha }$ contacts (grey dashed lines) encoded in the contact map $\mathbf{C}=\left\{c_{ij}\right\}$, where $c_{ij}=1$ (black) represents a contact and 0 represents non-contact (white). (b) The probability density $P$ (increasing from dark to light) plotted as a function of the accuracy of the backbone conformation LDDT and contact map similarity $\phi \left({\mathbf{C}}_n,{\mathbf{C}}_p\right)$ between the native fold ${\mathbf{C}}_n$ and model ${\mathbf{C}}_p$. (c) Similar schematic of a folded protein to that in (a) with the core residues colored grey and represented by the core residue identity vector $\mathbf{B}=\left\{b_i\right\}$, where $b_i=1$ (black) represents a core residue and 0 (white) represents a non-core residue. (d) Probability density $P$ (increasing from dark to light) plotted as a function of LDDT and the core identity similarity $\phi \left({\mathbf{B}}_n,{\mathbf{B}}_p\right)$ between the native fold ${\mathbf{B}}_n$ and model ${\mathbf{B}}_p$.