A molecular structure can contain thousands of atoms, but a useful question is often much smaller: which receptor residues are close to the ligand, which waters might matter, or where are the tight contacts? Selecting those nodes one by one is tedious—and small changes in distance or atom type can make it easy to miss part of the picture.
In SAMSON, the Node Specification Language (NSL) lets you describe these selections as expressions. A helpful way to build one is to think in layers: first identify the kind of nodes you want, then limit them to the relevant part of the structure, then add a distance condition. The examples below show how that approach can turn a broad structural question into a targeted selection.
Start with the biological roles
Many contact questions begin by distinguishing the receptor from the ligand. For example, to find receptor residues within 6 Å of a ligand while excluding glycine and alanine, use:
((n.t r and not r.t GLY,ALA) in n.c rec) w 6A of n.c lig
This expression combines several filters: it specifies residues (n.t r), excludes two residue types, keeps nodes in the receptor category, and applies a 6 Å proximity condition relative to the ligand. This is useful when a distance shell alone would include nodes you do not want to inspect.
If the question is about atoms rather than residues, make that distinction explicit. To select ligand heavy atoms for contact analysis:
(n.t a in n.c lig) and not a.s H
That selection is a useful starting point for atom-level checks, such as identifying close contacts or examining ligand geometry. A related expression identifies ligand atoms within 2 Å of receptor atoms:
(n.t a in n.c lig) w 2A of n.c rec
Use distance to focus, and exclusion to refine
The examples use w for a within-distance condition and b for a beyond-distance condition. This makes it possible to distinguish a contact shell from a region outside it. For instance, receptor atoms within 5 Å of a ligand but more than 3 Å from water oxygen atoms can be selected with:
(n.t a in n.c rec) and (n.t a w 5A of n.c lig) and (n.t a b 3A of (a.s O in n.c wat))
The value of this pattern is that it expresses two different spatial questions in one selection: “near the ligand” and “beyond the specified distance from water oxygens.” It can help focus an inspection on ligand-facing receptor atoms without including those close to water oxygen atoms.
You can also target specific interaction candidates. To find backbone nitrogen atoms within 3 Å of a ligand:
(a.s N in n.t bb) w 3A of n.c lig
To find water oxygens positioned within 3 Å of both receptor and ligand, combine the two proximity conditions:
(a.s O in n.c wat) and (n.t a w 3A of n.c rec) and (n.t a w 3A of n.c lig)
These are candidate selections, not conclusions about whether an interaction is energetically favorable. They help narrow the structure to nodes worth examining further.
Make selections answer a precise question
NSL can combine spatial filters with residue properties. For aromatic residues near a ligand, for example:
(r.t PHE,TYR,TRP,HIS) and (n.t r w 5A of n.c lig)
To omit histidine from that aromatic-residue selection, use:
r.t PHE,TYR,TRP w 5A of n.c lig
Likewise, a binding-site selection can be limited to a particular chain or residue range. To select atoms in residues 30–50 of chain A:
n.t a in (r.id 30:50 in c.n "A")
Combining a spatial neighborhood with a chain or residue filter is especially helpful when a structure contains multiple chains, ligands, waters, or other components. It keeps the expression aligned with the biological question instead of relying on visual selection alone.
A practical way to build and check an expression
- Write the target in plain language. Decide whether you need atoms, residues, waters, or another node type.
- Choose the structural context. Use categories such as receptor or ligand, or narrow by chain and residue identifiers.
- Add one distance condition. Begin with a single cutoff, then add exclusions only when they answer a specific question.
- Check what the expression selects. A selection is a filter for inspection; it does not by itself establish a chemical interaction or its strength.
This stepwise approach makes long expressions easier to read and adjust. It also makes the selection logic easier to share with a colleague: the expression records not only the cutoff, but the types of nodes and structural context being considered.
For more selection patterns and NSL attributes, see the NSL examples reference.
Note: SAMSON and all SAMSON Extensions are free for non-commercial use. Get SAMSON.