Figure 1

Model intrachain attractive potentials. (a) An example LRIM (black) and the corresponding SRIM (solid red) and modified SRIM (dashed red) potentials. (b) Comparing our native-centric modified SRIM potential (dashed red) with that in the desolvation-barrier (db) model (blue) [15] and the nonnative-centric SRSW model of Taylor et al. with $\lambda =1.05$ (black) [21].Reuse & Permissions

Figure 2

Comparison of experimental logarithmic folding rates with their simulated counterparts in the (a) SRIM and (b) LRIM. All $k_f^{\mathrm{sim}}$ values were determined using 500 trajectories except 104–180 trajectories were used for the slowest folding 1aps, 1fnf, 2qjl, and 1fkb. The lines are the best linear fits with correlation coefficients $r=0.83$ (a) and $r=0.82$ (b). Black, red, and blue data points are for all-$\alpha$, all-$\beta$, and $\alpha +\beta$ proteins, respectively. (Inset) Correlation between the simulated logarithmic folding rates in the SRIM model (${\log }_{10}{k}_f^{\mathrm{sim}}$) and modified SRIM model (${\log }_{10}{\tilde{k}}_f^{\mathrm{sim}}$) for a representative subset consisting of 4 all-$\alpha$, 6 all-$\beta$, and 5 $\alpha +\beta$ proteins specified in Table . The plotted best linear fit is given by ${\log }_{10}{\tilde{k}}_f^{\mathrm{sim}}=0.82563+1.2107·{\log }_{10}{k}_f^{\mathrm{sim}}$.Reuse & Permissions

Figure 3

(a) PDB structure and free energy profiles of Protein G (1pgb) in the LRIM (black), SRIM (red), modified SRIM (dashed red), and db (blue) models computed near each model's transition midpoint; $P\left(Q\right)$ is normalized conformational population. (b) Probability (right scale) maps of native contacts in the putative TSs of SRIM (top-left) and LRIM (bottom-right) which are marked, respectively, by the pink and gray regions in (a).Reuse & Permissions