Paste a peptide SMILES β get the one-letter sequence.
Free, instant, runs in your browser. Recognises all 20 natural amino acids
(A C D E F G H I K L M N P Q R S T V W Y),
flags side-chain-modified residues as K[x]
/ S[x]
/ Y[x] etc.,
identifies Aib explicitly, and handles cyclic peptides
(e.g. vasopressin β cyclic[CYFQNCPRG]).
What this tool does
Given a peptide drawn as SMILES (the chemical-structure string format), this converter
returns the canonical one-letter amino-acid sequence β the form you'd write in a paper or
an order. It works for linear peptides (e.g. leu-enkephalin β
YGGFL), for cyclic peptides (e.g. vasopressin),
and for modified peptide drugs (e.g. liraglutide, semaglutide) where
side-chain modifications such as fatty-acid lipidation on Lys produce a
K[x] flag in the output.
How it works
The converter parses your SMILES with RDKit MinimalLib running in your browser, then runs SMARTS substructure matches for each of the 20 natural amino-acid backbones, then walks the peptide bonds (atom-index continuity) to recover the NβC ordering. Side-chain modifications are detected by splitting each modifiable residue into "free" and "modified" SMARTS variants. Nothing is uploaded β the entire computation runs on your machine.
What gets recognised
- All 20 natural amino acids, displayed as single letters.
- Side-chain-modified naturals (phospho-Ser/Thr/Tyr, glyco-Ser/Thr, lipidated Lys, S-alkyl Cys, β¦) β
letter[x]. - Aib (Ξ±-aminoisobutyric acid) β
[Aib]. - Cyclic peptides (homo-detic head-to-tail or via Cys disulfide) β
cyclic[β¦]. - Multiple chains separated by
/if the peptide-bond graph breaks at unrecognised residues. - Stereochemistry is ignored (D vs L not distinguished).
Privacy
Your SMILES never leaves your browser. The conversion happens locally via RDKit compiled to WebAssembly. This page makes no network request once RDKit has loaded, so it's safe to use for proprietary structures.