Realistic Transition Paths for Large Biomolecular Systems: A Langevin Bridge Approach

TL;DR

The paper develops SIDE, a Langevin-bridge–based framework for generating realistic transition paths between protein conformations using a Go-like coarse-grained backbone potential coupled with a Rouse elastic network. It systematically compares SIDE to MinActionPath and EBDIMS across multiple proteins, showing that SIDE delivers smooth, low-energy trajectories that preserve backbone geometry and often reveal experimental intermediates. The authors also analyze the method’s limitations via the VATPase rotor case, underscoring the inherent trade-offs of coarse-grained models. Overall, SIDE offers a computationally efficient approach to exploring biomolecular conformational transitions and can be extended with more detailed coarse-grained representations to broaden applicability.

Abstract

We introduce a computational framework for generating realistic transition paths between distinct conformations of large bio-molecular systems. The method is built on a stochastic integro-differential formulation derived from the Langevin bridge formalism, which constrains molecular trajectories to reach a prescribed final state within a finite time and yields an efficient low-temperature approximation of the exact bridge equation. To obtain physically meaningful protein transitions, we couple this formulation to a new coarse-grained potential combining a Go-like term that preserves native backbone geometry with a Rouse-type elastic energy term from polymer physics; we refer to the resulting approach as SIDE. We evaluate SIDE on several proteins undergoing large-scale conformational changes and compare its performance with established methods such as MinActionPath and EBDIMS. SIDE generates smooth, low-energy trajectories that maintain molecular geometry and frequently recover experimentally supported intermediate states. Although challenges remain for highly complex motions-largely due to the simplified coarse-grained potential-our results demonstrate that SIDE offers a powerful and computationally efficient strategy for modeling bio-molecular conformational transitions.

Figures (12)

• Figure 1: The Fermi like function $g(x)$ (shown for $d_0=14$ and $a_0=1$).
• Figure 2: Conformational transitions in Adenylate kinase (AKE. (A) the open state (PDB code: 4AKE); and (B) the closed state (PDB code: 1AKE). AKE consists of three well-defined domains, the rigid CORE (blue, residues 1–29, 60–121, and 160–214) the nucleotide triphosphate binding domain, LID (green, residues 122–159), and the nucleotide monophosphate binding domain NMP (red, residues 30–59). The angle LID-CORE $\theta_1$ is formed by the centers of mass of the backbone of residues of LID ( residues 123–155 (LID), hinge (residues 161–165), and CORE (residues 1–8, 79–85, 104–110, and 190–198), whereas the angle NMP-CORE $\theta_2$ is formed by the centers of mass of the backbone of residues of NMP (residues 50–59), CORE ( residues 1–8, 79–85, 104–110, and 190–198), and hinge (residues 161–165).
• Figure 3: SIDE trajectory between the open state and closed state of AKE. The protein is represented as a string of beads, corresponding to the C$_{\alpha}$ atoms along its main chain, and colored based on its domain definition (see Figure \ref{['fig:ake']} for details). The elastic network at each pose is illustrated with edges colored in salmon pink. The timing for each pose is given as a fraction of the total time given for the construction of the path.
• Figure 4: Four different trajectories between the open state and closed state of AKE. We computed trajectories between the open state (1ake) and the closed state (4ake) using SIDE (black), CLD (red), EBDIMS (magenta) and MAP (blue). (A) The angle between the AMP binding domain (NMP) and the CORE is plotted against the angle between the LID and the CORE. (B) Projections of the 4 trajectories along the 2 principal components of their conformational spaces.
• Figure 5: The distributions of the distances between neighboring C$_\alpha$s for one snapshot along the trajectory between the open state and closed state of AKE are shown as violin plot. For each method, we picked the snapshot whose corresponding distribution has the largest variance. Note the presence of a peak in those distribution at the C$_\alpha$-C$_\alpha$ distance of $2.9 \AA$ corresponding to the cis Proline 87.