Two structures can show different conformations of the same protein, but that does not automatically tell you how the molecule moves between them. A straight coordinate blend may be easy to imagine, yet it can distort local geometry. For molecular modelers, the practical challenge is to create and inspect a continuous set of intermediate structures without mistaking a geometric interpolation for a proven physical mechanism.
SAMSON’s ARAP (As-Rigid-As-Possible) Interpolator provides one way to build that intermediate path. It uses a geometric model to generate conformations between a start and goal structure. The result can help visualize a transition, prepare a path for further analysis, or provide input for simulation workflows. It should be treated as a candidate path—not, by itself, evidence that the protein follows that route or that every intermediate is energetically favorable.
Start with comparable structures
The example in the tutorial uses two conformations of diphtheria toxin, PDB structures 1DDT and 1MDT, and works with chain A. The important preparation lesson is broader than this example: make sure the structures you intend to compare represent the corresponding molecular system. Remove chains that are not part of the comparison, then prepare both structures to remove alternate locations, water, ligands, and ions. Besides keeping the comparison focused, this helps avoid disconnected components that can prevent the calculation from running.
In SAMSON, the tutorial fetches both PDB entries, deletes chain B from 1MDT, and runs Home > Prepare. It then creates a named conformation for each cleaned structure. These saved conformations become the explicit start and goal states for interpolation.
Atom correspondence and connectivity matter
An interpolation needs to know which atoms in the start structure correspond to atoms in the goal. For the documented example, choose All except hydrogens under matching atoms. This uses non-hydrogen atoms for matching and as ARAP vertices. Other available choices include all atoms, only α-carbons, or only backbone atoms. The right selection depends on what you want the path to represent; for example, a backbone-focused representation omits side-chain atoms from the matching set.
Next, choose how the ARAP vertices are connected by edges. For the example, enable from bonds in the Start structure and Try connecting α-carbons before and after missing residue segments. The first uses covalent bonds in the start structure; the second adds connections across gaps in the residue sequence where residues are missing. These settings define geometric connectivity for interpolation, so they are not merely display preferences. The app also offers other edge-building rules, such as consecutive α-carbons or distance-based pairs.
Build a path you can inspect
Before running the calculation, enable Perform alignment before interpolation so the start conformation is aligned with the goal. Set the number of path conformations to 20; this count includes both endpoints. Once the calculation finishes, use the app’s slider to move through the intermediates rather than judging the transition from its endpoints alone.
Inspect the motion for abrupt changes and look at the generated ARAP edges in the Document view to understand how connectivity was constructed. If the app reports that the structure does not make one connected component, check for leftover water, ions, ligands, or other disconnected parts, then prepare both structures again.
Keep the next step in view
You can export the result as a path object in the current SAMSON document or export the path and its conformations to PDB files. A generated path can be used as a starting point for subsequent work, such as umbrella-sampling setup or path refinement; additional analysis is needed to assess its suitability for a particular scientific question.
For the full walkthrough and details of the available options, see the ARAP protein interpolation tutorial.
Note: SAMSON and all SAMSON Extensions are free for non-commercial use. Get SAMSON.