Two structures of the same protein can show different conformations, but the gap between them is often the difficult part. A simple visual interpolation may not reveal a plausible route, and a list of coordinates does not tell you how the protein might move from one state to another. Protein Path Finder in SAMSON offers a way to search for candidate transition paths and inspect the resulting motions.
The key is to treat the output as a set of paths to explore—not as a definitive account of how the protein moves. The workflow combines T-RRT pathway search with ARAP motion generation, using constrained minimization during the search to help keep generated states physically reasonable. The practical outcome is a collection of candidate conformations that you can review, compare, and export.
Start with a comparable pair of structures
Before searching, make sure the start and goal conformations describe the same system: matching chains, residues, and atoms. If your conformations come from separate files, prepare them individually and combine them as separate models in one PDB file. For your own structure, the documented preparation steps include removing alternate locations, ligands, solvent, and ions, and adding hydrogens. If residues or heavy atoms are missing, the PDBFixer extension can help repair the protein.
The tutorial’s sample document contains two conformations of adenylate kinase, corresponding to structures 4AKE and 1AKE. Starting with this sample is useful if you are new to the workflow because the models are already prepared and placed in one document.
Choose what is allowed to guide the motion
In the Protein Path Finder settings, select an interaction model and state updater. The documented example uses Universal Force Field (UFF) and FIRE, and the system should be minimized before the search. The FIRE updater is configured in the tutorial with a 1 fs step size and one step; use the tutorial’s settings as an example workflow, not as a universal prescription for every protein or project.
Next, choose the start and goal conformations from the active document. Then define the active ARAP atoms: these are the atoms that control the motion, while the remaining atoms follow according to the ARAP method. In the adenylate kinase example, two backbone alpha carbons—CA in GLY 12 and CA in ARG 123—are selected. This is a small, specific choice, not a rule that those residues should guide every protein transition. The atoms you designate and the sampling region influence the motion the search can explore.
Think of search settings as exploration controls
The sampling box defines where active atoms may be sampled. Its size and position bias their motion, so review the box rather than accepting it without inspection. The tutorial sets a 200-angstrom cube for its example system; that value is part of the example setup, not a general recommendation.
Search parameters also affect the exploration. Runs sets how many searches are performed, and the tutorial uses two runs to seek up to two paths. A fixed seed makes a run reproducible under the same setup; without it, the app uses a random seed. Other controls include ARAP-modeling iterations, constrained-minimization iterations, temperature settings for T-RRT, extension step size, alignment strategy, and the maximum time per run. These settings govern the search process, so record them alongside results when you want to compare or reproduce a run.
Review paths, not just the final structure
When a path is found, select it in the Results tab and use the slider to move through its conformations. The app displays the corresponding structure in the viewport and its energy in the UFF window. You can also compare energy curves for multiple selected paths. The table reports values such as minimum and maximum energy, saddle, barrier, and the first and last energies.
Those values are useful for comparing the computed paths within the chosen setup, but they should not be mistaken for experimentally established transition barriers. They depend on the model and calculation used. Review the motion itself, the energy profile, and the search conditions together. If a path is useful, export it as a trajectory or export selected conformations at an interval. The tutorial also describes refining resulting paths with the P-NEB app.
This workflow helps turn two endpoint structures into inspectable candidate motions, while keeping important choices—model preparation, active atoms, sampling region, and search settings—visible and reproducible.
For the full procedure and interface details, see the Protein Path Finder tutorial.
Note: SAMSON and all SAMSON Extensions are free for non-commercial use. Get SAMSON.