Fig. 4: Learning the mechanism of polymer folding.
From: Machine-guided path sampling to discover mechanisms of molecular self-organization
a, Representation of the learned mechanism. The heat map (color bar) represents a reduced explicit model of the committor pB = pF to the folded state as reproduced by the expression \({q}_{{\mathrm{B}}}\left(U,{Q}_{6}\right)=\alpha (U-{U}_{0})+\beta \log \left({Q}_{6}-{Q}_{6,0}\right)+\gamma\), where U is the total potential energy of the polymer, Q6 quantifies its crystallinity, and the numerical constants are α = −7.144, β = 3.269, γ = 11.942, U0 = −2.351 and Q6,0 = 0.035. Insets: molecular configurations of the polymer at pB = 0, 0.5 and 1. Polymer beads are colored according to their value of q6, from white (low values) to dark orange (high values). b, Validation of the learned committor. Cross-correlation between the committor predicted by the trained network and the committor obtained by repeated sampling from molecular configurations on which the algorithm did not train. The average of the sampled committors (blue line) and their s.d. (orange shaded) were calculated in bins of the learned committor indicated by the vertical steps. For reference, the red line indicates the identity.
