WITH abYcloud
8× FASTERTarget → validated candidate
Founded in London. Connecting AI design, automated wet-lab testing and cryo-EM to advance antibody programmes in the UK and beyond.
How it works
Each campaign sharpens the next. Generative design feeds the auto-lab; the auto-lab feeds cryo-EM; cryo-EM feeds the model.
Evidence
Target → validated candidate
Per optimisation round
Cryo-EM cost cut
Cryo-EM structure cost
London · United Kingdom
abYcloud began with three UCL researchers and a shared question: how can we make antibody discovery more rigorous and accessible? Our UK company connects methods research in London with automated wet-lab and structural validation capabilities in Shanghai and Suzhou, and international partnerships in Hong Kong. For UK biotech, pharma and academic teams, that means one connected workflow from a target to experimentally informed antibody candidates.
“Trade the execution, not just the idea.”
An embedded antibody-engineering function for UK biotech: sequence mining, structure-informed design and experimental feedback, with optimisation rounds reported at roughly three weeks across multiple campaigns.
Academic connections at UCL, Imperial College London and Xi’an Jiaotong-Liverpool University, alongside a clinical-translational pilot with a major Hong Kong university hospital.
Featured by the UCL Centre for Digital Innovation and selected for the 2025 P4 Precision Medicine Accelerator cohort, connecting research with therapeutic development.
UK research, connected to experimental evidence
London methods research connects to our automated lab and structural biology network. Antibody designs are tested, measured and returned to the model so each programme can build on experimental evidence.
One loopfrom computational design to experimental feedback, across our international team.
Network · London ↔ Shanghai ↔ Hong Kong
Built on peer-reviewed research
Do antibody CDR loops change conformation upon binding?
mAbs · 2024
AsEP: Benchmarking Deep Learning Methods for Antibody-specific Epitope Prediction.
Advances in Neural Information Processing Systems 37 (NeurIPS 2024) — Datasets and Benchmarks Track · 2024
15 minutes, no setup. We'll show how the closed loop maps to your campaign.