For molecular modelers, one recurring challenge is precise atom or residue selection—especially in large or complex systems. Traditionally, users rely on syntax-heavy search commands, filters, or visual inspection, which can slow down workflows and introduce errors. However, there’s a more intuitive path available in SAMSON: using natural language selection via /select commands in SAMSON AI.
Instead of scrolling through node trees or learning new selection syntaxes, modelers can now simply describe what they want using everyday language. Imagine typing: /select all lysines in chain B or /select all side chains within 5 angstrom of the ligand. The assistant interprets and converts these phrases into precise queries using SAMSON’s Node Specification Language (NSL).
When is this helpful?
This can be particularly valuable when:
- Working with unfamiliar structures or databases
- Needing fast, iterative selections for tasks like binding site analysis
- Teaching others molecular modeling without diving into query syntax
Even better: SAMSON AI shows the actual NSL expression it’s executing in response to your natural language, which turns every interaction into a mini training opportunity.
How it works
Start by opening SAMSON AI via Interface > Assistant or the shortcut Ctrl+0 (Windows/Linux) or Cmd+0 (Mac).
Then, simply type a query like:
/select the binding site/select all lysines in chain B/select all side chains within 5 angstrom of the ligand
SAMSON AI will:
- Parse your natural language input
- Translate it into a structured NSL query
- Execute it and highlight the selected elements in your document
Example in Action
Here is the Assistant’s command list interface, which pops up when you start typing /:
Why this matters
Time spent learning or navigating selection syntax adds up—especially with large molecular systems or when switching between software tools with different standards. With SAMSON AI, you’re focusing on your goal (“select side chains near the ligand”) rather than on the how.
This helps modelers of all experience levels stay productive, while also learning the underlying query language by example. It also ensures consistency and repeatability in a way that visual inspection can’t guarantee.
If you work in structure-based drug design, protein engineering, or molecular dynamics, this can speed up your modeling sessions and reduce mistakes linked to manual or ambiguous selections.
To explore more examples and commands that integrate text-based inputs with SAMSON’s modeling engine, visit the full documentation:
https://documentation.samson-connect.net/users/latest/samson-ai/
SAMSON and all SAMSON Extensions are free for non-commercial use. You can download SAMSON at https://www.samson-connect.net.