A molecular modeler may have two structures of the same protein and still face a difficult question: what motion could connect them? The endpoints are known, but the route between them is not. Simply switching from one structure to the other does not reveal the intermediate conformations—or whether a proposed motion is worth examining.
SAMSON’s Protein Path Finder is designed to search for candidate transition paths between two protein conformations. It uses ART-RRT: T-RRT for pathway search and ARAP for generating motion, with constrained minimization during the search to help keep generated states reasonable. The result is a set of paths to inspect, not a claim that the protein must follow one unique route.
Start with compatible endpoints
The first practical challenge is often preparing the input. The start and goal structures need to describe the same system: the chains, residues, and atoms should match. If the conformations come from separate files, prepare them individually, then combine them in one PDB file as separate MODEL entries. For a new workflow, the tutorial’s sample document is a useful way to learn the controls before adapting them to your own protein.
For your own model, SAMSON’s preparation tools can help remove alternate locations, ligands, solvent, and ions, and add hydrogens. If residues or heavy atoms are missing, the PDBFixer extension can be used to repair the structure. The tutorial’s sample is already prepared and contains two conformations of adenylate kinase: 4AKE as the start and 1AKE as the goal.
Choose what guides the motion
Protein Path Finder uses selected active atoms to control the motion; other atoms follow them through the ARAP method. In the sample, two alpha-carbon atoms are chosen: one in GLY 12 and one in ARG 123. These are examples for that document, not universal choices. Think about which parts of your protein should guide the transition, and inspect the selection before starting a search.
The app also uses a sampling box to define the region in which active atoms can be sampled. Its size and position influence the search and therefore the paths that may be found. The tutorial demonstrates a 200 Å cube. Treat that as an example setting, not a general rule: the useful region depends on the motion you want to explore. A box that is too restrictive can limit possible movement, while an unnecessarily large region may allow sampling far from the motion of interest.
Set up energy evaluation before searching
The tutorial uses the Universal Force Field (UFF) as the interaction model and FIRE as the state updater. The system should already be minimized; SAMSON’s Edit > Minimize command uses UFF. When prompted during setup, choose to use existing bonds. The sample workflow sets the FIRE step size to 1 fs and the number of steps to 1.
These choices make the example reproducible as a workflow, but they are not a substitute for judging whether a model and energy setup are appropriate for a particular scientific question. The minimized states and resulting paths should be reviewed in context.
Make the search interpretable
For the tutorial run, the settings include two runs, 20 ARAP-modeling iterations, 20 constrained minimization iterations per generated state, and an RRT extension step size of 1 Å. The T-RRT settings shown are an initial temperature of 0.001 K, a temperature factor of 2, and one failure before increasing the temperature. A maximum elapsed time per run can also be set.
If you want to repeat a search, enable Use seed and record the seed value. The seed is incremented after each run; with the option unchecked, a random seed is used. Reproducibility is especially helpful when comparing changes to active atoms, the sampling box, or search parameters. Keep a record of those choices alongside the seed, rather than relying on memory.
During a run, the Advanced information panel reports elapsed time, the number of tree nodes, the run number, and how many paths have been found. You can pause and resume the planner, or stop it. These indicators help you follow progress, but a large node count or long run time is not itself evidence that a path is biologically plausible.
Inspect paths, not just the endpoint
Found paths appear in the Results tab. Select a path to view its energy curve, then use the slider to inspect conformations along the route. The table reports quantities such as the minimum and maximum energy, the energy difference called Saddle, and the Barrier relative to the first conformation. These summaries can help compare candidates, but reviewing the structures and energy profile together is more informative than choosing a path from a single number.
You can export paths as trajectories into the document, export sampled conformations, or copy table data and path energies for further analysis. The resulting paths can also be refined with the P-NEB app; refinement is a useful next step when a candidate route warrants closer examination.
For the complete setup instructions and interface details, read the Protein Path Finder tutorial.
Note: SAMSON and all SAMSON Extensions are free for non-commercial use. Get SAMSON.