A crowded molecular scene can make a simple question surprisingly time-consuming: which model should you inspect first? Is one structure much larger than the others? Does it contain several chains, many residues, or a charge that makes it worth checking? Opening models one at a time is a slow way to answer questions that can often be expressed as filters.
In SAMSON, the Node Specification Language (NSL) lets you describe structural models by their attributes. The structuralModel attribute space has the short name sm. Its attributes include model-level counts, total formal and partial charge, and inherited properties such as name and visibility. That makes it useful for an initial inventory: narrow the set of models by a measurable property, then inspect the candidates that remain.
Start with size and composition
If you need to find models above a rough size threshold, try sm.nat > 1000. Here, nat is the short name for numberOfAtoms. To look for models in a range, use a range expression such as sm.nat 100:200. The same pattern applies to other integer counts, including chains, molecules, residues, and segments.
For example, sm.nc 2:4 filters by number of chains, while sm.nr 100:130 filters by number of residues. These are useful first-pass questions when a project contains structures at different levels of complexity. They do not tell you whether a model is biologically relevant or correctly prepared; they help you find models whose basic size or organization matches what you want to review.
Composition counts can provide another quick clue. Use sm.nC for carbon atoms, sm.nH for hydrogen atoms, sm.nN for nitrogen atoms, sm.nO for oxygen atoms, and sm.nS for sulfur atoms. For instance, sm.nO 10:20 selects models with an oxygen count in that range. The attribute list also includes coarse-grained atoms, structural groups, and formal charge.
Use charge as a screening clue
Structural models expose both formalCharge (sm.fc) and partialCharge (sm.pc). They answer different questions: formal charge is an integer, while partial charge can be a real number. Examples from the NSL reference include sm.fc 6:8 and sm.pc 1.5:2.0.
These filters can help you locate models for closer inspection, but a matching total charge does not establish that a structure is chemically correct. Treat it as an inventory or review aid, not a substitute for checking the model and its preparation.
Make the first pass manageable
A practical workflow is to begin with one clear question: “Which models have fewer than 1,000 atoms?” or “Which models have two to four chains?” Use the corresponding attribute expression to narrow the scene, then inspect the resulting candidates. If you are trying to locate a named model, the inherited name attribute also supports expressions such as sm.n "A" and sm.n "L*" in the reference.
Visibility and selection attributes can help describe scene state too. For example, sm.v refers to visible structural models and not sm.selected refers to models that are not selected. These inherited node attributes are handy when the problem is not the model’s composition, but simply finding the relevant items in a busy workspace.
The key is to use structural-model attributes for structural-model questions: counts describe the model as a whole, rather than asking you to inspect every atom individually. Start with a broad, interpretable filter, and use the result to guide your next inspection step.
For the complete attribute list and additional examples, see the structural model attributes reference.
Note: SAMSON and all SAMSON Extensions are free for non-commercial use. Get SAMSON.
