A molecular project can quickly accumulate structural models that are difficult to compare by eye. One may be a small ligand complex, another a large assembly, and a third may have a different charge or chain count than expected. Opening every model to inspect it takes time—and makes it easy to overlook a useful candidate.
SAMSON’s Node Specification Language (NSL) offers a practical first pass: filter structural models using properties such as atom count, number of chains, residue count, elemental composition, and charge. The key is to treat these as triage questions. A filter can narrow the list; it does not, by itself, establish that a model is biologically correct.
Start with the size of the model
When a document contains models at very different scales, atom count is a useful way to find structures that merit inspection. The sm.nat short name refers to numberOfAtoms. For example, sm.nat < 1000 matches structural models with fewer than 1,000 atoms, while sm.nat 100:200 selects models in the documented range from 100 to 200 atoms.
This can help separate small structures from larger ones before you look more closely. It is also useful when you know roughly how large a candidate should be but do not remember its name. Think of the result as a shortlist, not a verdict: atom count alone says nothing about whether the atoms are arranged correctly.
Use biological organization as another clue
Size does not tell the whole story. A model’s chain and residue counts can help distinguish structures with different organization. Use sm.nc for the number of chains and sm.nr for the number of residues. For instance, sm.nc 2:4 searches for models with a chain count between 2 and 4; sm.nr > 100 finds models with more than 100 residues.
These filters are handy when checking a collection against a rough expectation—for example, finding models with multiple chains or focusing on structures with a substantial residue count. They can also reveal unexpected entries: if a model you thought was a single-chain structure appears in a multi-chain search, it may be worth checking what was imported or included in the model.
Check composition and charge
Element counts provide another way to narrow candidates. NSL includes short names for carbon, hydrogen, nitrogen, oxygen, and sulfur counts: sm.nC, sm.nH, sm.nN, sm.nO, and sm.nS. The reference page gives examples such as sm.nO 10:20 and sm.nS < 10. These counts can help distinguish models with different compositions, but they should not be mistaken for a chemical identity check.
Charge filters can be useful when the expected charge is known. sm.fc searches by total formal charge, using integer values; sm.pc searches by total partial charge, using real values. For example, sm.fc 6:8 selects models whose formal charge falls in that range, and sm.pc 1.5:2.0 searches by partial charge. Formal and partial charge are different attributes, so choose the one that matches the question you are asking.
A simple, repeatable triage routine
- Ask one question at a time. Begin with a property you know, such as a maximum atom count or an expected chain-count range.
- Use the short names. Expressions such as
sm.nat,sm.nc, andsm.fcare compact and shown in the reference examples. - Check the resulting models. A match means the model satisfies the specified attribute, not that it meets every scientific requirement.
- Follow up with a second property. If the first result is still broad, search again using another attribute, such as residue count or oxygen count.
Keeping filters focused makes the process easier to understand and troubleshoot. It also makes your selection criteria easier to communicate to a collaborator: you can say which model property you used, rather than relying on a visual impression or an ambiguous filename.
For the complete attribute list, syntax examples, and details on inherited attributes such as visibility and selection, see the SAMSON structural model attributes reference.
Note: SAMSON and all SAMSON Extensions are free for non-commercial use. Get SAMSON.
