Many molecular modelers face a recurring challenge when studying functional protein dynamics: how to connect two known conformations of a protein—say, an open and a closed state—with a realistic transition path. This is a critical step for gaining insights into structural mechanisms, validating simulations, or preparing data for enhanced sampling or machine learning workflows.
If you have start and goal conformations of a protein and want to find a feasible transition between them without resorting to computationally expensive, full-blown molecular dynamics simulations, the Protein Path Finder app in SAMSON could provide the solution.
In this blog post, we will walk you through a core part of using the Protein Path Finder: setting up the system and defining the search parameters to find conformational paths using the ART-RRT method, which combines Rapidly-exploring Random Trees with As-Rigid-As-Possible modeling.
System Setup: Start and Goal Conformations
After loading your input model (either your own or the one from the tutorial sample), begin by identifying the start and goal conformations of your protein. The app lets you detect these directly from the current SAMSON document, assuming you’ve loaded the system with multiple models. For example, in the case of Adenylate Kinase, the two conformations correspond to the PDB structures 4AKE and 1AKE.
Selecting Active Atoms for ARAP
The As-Rigid-As-Possible method requires a small set of active atoms that guide the conformational transition. Typically, these are backbone atoms from key hinge residues. In our example, alpha carbons from residues GLY 12 and ARG 123 are selected.
SAMSON makes this easy: a pre-created group in the document already includes them. Double-click on the group to highlight the atoms, then assign them as active ARAP atoms in the app.
Defining the Sampling Box
The sampling box constrains the region in which the active atoms can move to generate candidate paths. For most proteins, a cubic box that tightly encompasses both conformations works well. In this tutorial, a 200Å cube is used.
Choosing Search Parameters
Customizable search parameters let you control how aggressively or conservatively the space is explored. For example:
- Runs: How many independent paths you want to find (e.g., 2).
- ARAP-modeling iterations: Typically 20.
- Minimization iterations (FIRE): 20-step constrained minimization helps refine each conformation.
- Initial Temperature and Sampling Strategy: These influence the behavior of the T-RRT algorithm.
Initiating the Search
Once the above configuration is complete, click Run to start computing transition paths. You can monitor the number of nodes explored, runtime, and how many paths were found in real-time.
These found paths can then be further analyzed, export trajectories, or refined using Nudged Elastic Band methods for improved energetic accuracy.
Setting up a system correctly for conformational path finding is often an overlooked yet crucial task. Understanding which atoms to activate, how to define the sampling box, and how to tune search parameters gives you full control over this process—helping you produce meaningful, reproducible results without unnecessary computational costs.
To go deeper, including results visualization and exporting paths, refer to the full Protein Path Finder tutorial.
SAMSON and all SAMSON Extensions are free for non-commercial use. You can download SAMSON at https://www.samson-connect.net.