Glycomics statistics
Differential abundance, motif enrichment and time-series analysis for N- and O-glycomics, built on compositional data analysis so results hold up to scrutiny.
Services
From raw spectra to a decision
Six areas where glycoscience questions stall most teams, and where our methods move them forward.
Structures from MS/MS
Deep-learning annotation of tandem mass spectra into full glycan structures with CandyCrunch, including diagnostic ion analysis for difficult isomers.
Binding and specificity
Predict and interpret how lectins, antibodies and microbial adhesins recognize glycans, from array data to target-level specificity maps.
Biosynthetic modeling
Trace a glycome shift back to the enzymes and pathways that drive it, to guide cell line engineering or explain a disease phenotype.
3D structure at scale
Glycan conformations and protein-glycan contacts analyzed across thousands of structures with GlyContact, for glycoprotein design and epitope work.
Custom models and data
Machine learning models, curated datasets and pipelines built for your glycan or glycoprotein question, delivered as code your team owns.
Engagements
Ways to work together
Expert consultation
Focused sessions on a specific question: data review, method choice or a second opinion before a decision.
Scientific advisory
A standing monthly arrangement for companies that need regular glycoscience input. Equity can form part of it.
Defined projects
An agreed analysis with clear deliverables: a report, publication-quality figures and reproducible code.
Open methods
Built on tools the field uses
The software behind our work is open source and used by glycobiology labs and companies worldwide. You get the people who wrote it.
- glycoworkThe Python toolkit for glycan data science: parsing, statistics, motifs, networks and GlycoDraw figures like the ones on this page.
- CandyCrunchDeep learning that predicts glycan structures directly from LC-MS/MS data.
- GlyContactGlycan 3D structure and protein contact analysis at scale.
- LectinOracleDeep learning model for protein-glycan binding prediction.
About
Daniel Bojar, PhD
Founder and principal consultantAssociate Professor of Bioinformatics at the University of Gothenburg and Deputy Director of the Centre for Modelling Life at the Membrane with AI (CEMERAI). Trained at ETH Zurich and at MIT and the Wyss Institute at Harvard. His lab develops widely used machine learning methods for glycobiology.
Current and past clients include pharmaceutical companies and venture-backed biotech in Europe and the US.
- 2027David Y. Gin New Investigator Award, ACS CARB
- 2026Young Researcher Award, International Carbohydrate Organization
- 2026Roland Schauer Early Career Researcher Award, GBM
- 2022Forbes 30 Under 30 Europe
53peer-reviewed publications and preprints
75MSEK in research funding as main applicant
17invited conference talks since 2021
ERCStarting Grant holder