What is SMILES to SDF?
SMILES to SDF converts SMILES strings into three-dimensional SDF (Structure Data File) format using RDKit. SDF is the standard input format for most docking and cheminformatics workflows — while SMILES encodes connectivity, SDF carries 3D atomic coordinates that downstream tools require.
How to use SMILES to SDF online
ProteinIQ runs the conversion server-side, generating 3D coordinates without requiring RDKit or any other software to be installed locally.
Input
Paste one SMILES per line into the text area, or use name–SMILES pairs separated by a tab (name<TAB>SMILES). Up to 10 entries are accepted per job. Alternatively, upload .txt, .csv, .tsv, .smi, or .smiles files, or fetch compounds directly from PubChem by name or CID.
| Format | Example |
|---|
| Bare SMILES | CC(=O)Oc1ccccc1C(=O)O |
| Named SMILES | aspirin\tCC(=O)Oc1ccccc1C(=O)O |
| File upload | .smi, .smiles, .csv, .tsv, .txt |
| PubChem fetch | Compound name or CID |
Settings
| Setting | Description |
|---|
Add hydrogens | Adds explicit hydrogens before 3D embedding (default: on). Required for accurate geometry and most docking pipelines. |
Optimize geometry | Runs a short UFF force field minimization after embedding (default: on). Improves coordinate quality for strained structures. |
Output
Each entry produces an individual .sdf file. A combined batch.sdf containing all successfully converted structures is also generated for workflows that accept multi-record SDF input.
How it works
RDKit's AllChem.EmbedMolecule places atoms in 3D space using distance geometry, seeding initial coordinates from a random conformation. When Optimize geometry is enabled, Universal Force Field (UFF) minimization refines bond lengths, angles, and torsions. The resulting conformation is a reasonable starting geometry but represents a single local energy minimum, not necessarily the global one.
Limitations
- Generates a single conformer per molecule. For ensemble docking or conformer sampling, generate multiple conformers separately.
- Very large or highly flexible macrocycles can fail embedding — RDKit distance geometry struggles with long-range constraint satisfaction in these cases.
- Tautomer and protonation state are taken as given from the input SMILES. Adjust ionization state before conversion if the target pH matters.
