---
title: File Formats & Conversion
description: >-
  Convert between SMILES, SDF, and PDB formats with RDKit for cheminformatics
  workflows.
---

Format conversion lets you move molecules between SMILES, SDF, and PDB so each stage of a workflow gets the representation it expects. Careful parsing — checking for `None`, sanitizing, and preserving provenance — is what most affects whether a batch runs clean or silently drops records.

## Parse molecules

```python
from rdkit import Chem

mol = Chem.MolFromSmiles("CC(=O)Oc1ccccc1C(=O)O")
if mol is None:
    raise ValueError("Invalid SMILES")

mol = Chem.AddHs(mol)
Chem.SanitizeMol(mol)
```

## SDF to SMILES

```python
from rdkit.Chem import SDMolSupplier

supplier = SDMolSupplier("input.sdf", removeHs=False)
smiles: list[str] = []
for mol in supplier:
    if mol is None:
        continue
    smiles.append(Chem.MolToSmiles(Chem.RemoveHs(mol)))
```

## SMILES to SDF

```python
from rdkit.Chem import SDWriter

writer = SDWriter("/workspace/outputs/library.sdf")
for s in smiles:
    mol = Chem.MolFromSmiles(s)
    if mol is None:
        continue
    Chem.SanitizeMol(mol)
    writer.write(Chem.AddHs(mol))
writer.close()
```

## PDB hygiene

- Use `Chem.MolFromPDBFile("structure.pdb", removeHs=False, sanitize=True)`.
- Strip waters/ions before minimization if they are not needed.
- Ensure residue/atom names are preserved before exporting back to PDB.

## Validation tips

- Always check for `None` after parsing; skip invalid records rather than
  crashing the batch.
- Normalize tautomers if your downstream model expects a canonical form.
- Record provenance: store the source file path in
  `mol.SetProp("source", path)`.