We build computational tools to read, analyze, and predict glycans: the complex, branched carbohydrates that coat every cell and shape infection, immunity, cancer, and development. Our software is open source, Python-first, and designed to work together, with glycowork as the shared foundation.
Lab website · X: @daniel_bojar
| Package | What it does | Install |
|---|---|---|
| glycowork | The hub: glycan sequence processing, motif analysis, glycomics statistics, biosynthetic networks, deep learning models, and GlycoDraw drawing. Docs |  |
| CandyCrunch | Predicts glycan structures from LC-MS/MS glycomics data with deep learning. |  |
| GlyContact | Retrieves, processes, and analyzes 3D glycan structures. Docs |  |
| GlycoForge | Simulates glycomics datasets with controllable biological effects, batch effects, and missing data. |  |
| GlycoGym | Benchmark suite of seven glycan property prediction tasks, plus the GlyVerse pre-training corpus. |  |
| glycorender | Renders GlycoDraw SVGs to publication-ready PDF and PNG. |  |
Draw any glycan in SNFG notation with one line of glycowork:
# pip install glycowork
from glycowork.motif.draw import GlycoDraw
GlycoDraw("Neu5Ac(a2-3)Gal(b1-4)[Fuc(a1-3)]GlcNAc", sticker=True) # sialyl Lewis X
All banner art on this page is drawn the same way; see scripts.
- Canonicalize: convert glycan sequences from any common notation to canonical IUPAC-condensed in the browser (code).
- GlycoGauntlet: an open competition to annotate LC-MS/MS glycomics data. Beat CandyCrunch or bring your expert annotations, and join the paper.
Analysis code and data accompanying our publications:
- Reaction-order: preferred reaction order from diamond motifs in biosynthetic networks
- FragmentFactory: human-interpretable diagnostic glycan fragments
- LectinOracle_MIMB: the LectinOracle chapter in Methods in Molecular Biology
- scGlycomics_b16_branching: predicting single-cell glycosylation features from scRNA-seq
Earlier models such as SweetNet, LectinOracle, and GIFFLAR are archived; their maintained implementations live in glycowork.
Computer labs for the course BIO513: glycomics, proteomics, and metabolomics.
Found a bug or missing feature? Open an issue in the relevant repository. Contributions are welcome, and new functionality usually fits best as an extension of glycowork or CandyCrunch rather than a standalone tool.