Short answer: yes. Glycans are a pain to encode in SMILES compared with amino acids/peptides. Proteins are basically obedient noodles: linear chains with one main bond type and mostly fixed stereochemistry. Glycans are origami hydras.
Why glycans are harder in SMILES:
- Branching everywhere. Oligosaccharides are highly branched trees, so a single glycan demands lots of SMILES ring/branch markers that explode in length and readability. Peptides are almost always linear.
- Many near-identical monomers. Swap one hydroxyl’s stereochemistry and glucose becomes galactose; add an N-acetyl and now it’s GlcNAc; every monosaccharide has multiple chiral centers to specify.
- Anomeric configuration and ring form. You must encode α/β at C1 plus whether the sugar is in a pyranose or furanose ring; peptides don’t have an extra “anomeric” headache.
- Linkage positions vary. Glycosidic bonds can be 1→2, 1→3, 1→4, 1→6, etc., and each branch’s linkage must be explicitly encoded; peptide bonds are basically one flavor.
- Real-world ambiguity. In glycomics you often only know a composition or a partially defined topology (e.g., positions or stereochemistry not fully resolved). SMILES requires a fully specified structure, which is why the field leans on carbohydrate-specific notations that tolerate ambiguity (WURCS, GlycoCT) and only convert to SMILES when everything is nailed down.
What people actually use:
- Dedicated glycan notations: WURCS (used by GlyTouCan) and GlycoCT exist precisely because general cheminformatics formats struggle with carbohydrate complexity and partial knowledge.
- Converters exist but need full detail: tools like GlyLES can translate IUPAC-condensed glycans into SMILES, which is handy for docking or substructure search, but only when the structure is completely specified.
- Tradeoffs: WURCS captures ambiguous or underspecified glycans better than SMILES, but it isn’t “chemically complete” for generic cheminformatics tooling, so teams sometimes juggle both.
Bottom line: peptides are easy mode; glycans come with branching, anomers, ring forms, linkages, and frequent ambiguity. If you need SMILES, generate it from a fully defined glycan; if your data are uncertain or compositional, stick to WURCS/GlycoCT and spare yourself the migraine.