From a Rough Molecular Route to a Better-Relaxed Transition Path

A molecular transition path can look plausible and still contain awkward intermediate structures: neighboring frames may be unevenly spaced, or the route may need relaxation under an interaction model. The key practical distinction is that path refinement is not path generation. The Parallel Nudged Elastic Band (P-NEB) app in SAMSON is useful when you already have a candidate route—or a sequence of conformations—and want to optimize its intermediate images.

That distinction can save time. If you only have two endpoint structures, first create intermediate conformations, for example by interpolation or with a path-generation tool. P-NEB needs that initial sequence to work from. It optimizes intermediate images while spring forces help keep neighboring states distributed along the route. It implements the climbing-image NEB method, which can also be used to search for a saddle point.

Start with a path when you can

SAMSON can apply P-NEB to either a path or a set of conformations. For the same system and number of steps, the tutorial recommends using a path because optimizing a set of separate conformations takes longer. If your input is a set, select the conformations in the Document view and use Conformation > Create path from conformations in the context menu to combine them into a path.

A path stores a trajectory for selected atoms; a conformation stores positions for selected atoms. This matters when choosing input: make sure the frames represent the same group of atoms and the intended progression between your endpoints. P-NEB improves the supplied route; it does not decide which biological event or pathway you meant to model.

A practical first run

1. Load a sample document or prepare your own candidate path, and add the P-NEB extension and FIRE state updater.
2. Open P-NEB from Home > Apps > All > P-NEB, or find it with the Find everything… search box.
3. For the tutorial’s example setup, enter a spring constant of 1.00 and 100 loops, choose Universal Force Field as the interaction model and FIRE as the optimizer, and check Parallel execution.
4. Leave Climbing image method unchecked for the first run. You can repeat the calculation with it enabled to try the climbing-image strategy.
5. Enter a suffix such as NEB, select the path in the Document view, and click Run.

When Universal Force Field setup asks whether to use existing bonds, the tutorial’s instruction is to choose use existing bonds and click OK. The app then initializes and runs the optimization; progress is visible in the status bar.

Review the output, not just the completion message

After optimization, SAMSON adds a new path to the Document view and displays a computation summary in the P-NEB interface. Select the resulting path and inspect its conformations; double-clicking a path starts or stops its animation. Reviewing the frames helps you understand what changed and whether the optimized route remains appropriate for your intended analysis. Optimization alone should not be treated as proof that a pathway is biologically correct or that it describes the full kinetics of a process.

If you are starting from two relaxed states, first generate a sequence between them—for example with linear interpolation or a path-generation extension such as Ligand Path Finder—then use P-NEB to refine that candidate. Once reviewed, the path can be exported or used as a starting point for a further simulation or free-energy workflow.

For setup details and the full walkthrough, read the original P-NEB tutorial.

Note: SAMSON and all SAMSON Extensions are free for non-commercial use. Get SAMSON.