Row SystemWave Shape Decorative svg added to top
Resource Library Rapid Cell-Free Validation of AI-Generated Antibody Libraries
Rapid Cell-Free Validation of AI-Generated Antibody Libraries Chris Icke, Ania Boczkaj, Sumit Kalsi, Andreas Waeber, Raman Pandey, Mozhdeh Khajvand, and Ruben Tomás Nuclera Ltd, Cambridge, UK.

Rapid Cell-Free Validation of AI-Generated Antibody Libraries in 14 Days

Artificial intelligence can generate thousands of promising antibody variants, but pushing these in silico libraries directly into traditional mammalian cell expression creates a massive validation bottleneck.

This poster showcases how to overcome these bottlenecks by inserting an automated, cell-free protein synthesis (CFPS) triage step upstream of mammalian biology. By rapidly triaging libraries with CFPS, computational teams can efficiently filter out AI-generated non-binders and reserve premium CHO validation strictly for confirmed hits.

In this poster, you'll discover:

  • A rapid 14-day workflow: How to take 185 digital sequences, express them, triage them, and convert them to decision-grade SPR data from 14 days.
  • Near-perfect expression success: How CFPS achieved a 98% build success for an anti-PD-1 library (92 of 94 variants) and a 100% build success for an anti-IL-6 library (94 of 94 variants), expressing them as intact full-length IgGs.
  • Massive downstream savings: How the binary triage screen successfully identified and diverted 139 non-binders away from costly CHO workflows. This included catching an 83% non-binder rate in the IL-6 library, preventing weeks of wasted mammalian expression and budget.
  • Reliable kinetic concordance: How cell-free SPR results yielded >91% concordance with mammalian CHO data (BLI and SPR), delivering decision-grade kinetics to support reliable early hit triage.

Why It Matters

Fast, predictive triage protects your mammalian pipeline. By replacing slow, unvetted CHO screening with a rapid cell-free binary triage and SPR workflow, discovery teams can stop wasted spend, trust their data, and focus expensive downstream resources exclusively on the right constructs.

Contributors

Chris Icke - Senior Scientist Workflow Development in Protein Science
Ania Boczkaj - Senior Scientist, Antibodies in Protein Science
Sumit Kalsi - Associate Director Product Research in DMF Research
Andreas Waeber - Senior Software Developer in Workflow Software
Raman Pandey - Principal Scientist in Protein Science
Mozhdeh Khajvand - Senior Workflow Engineer in Integration
Ruben Tomás - Sr Product Marketing Manager

Download: Rapid Cell-Free Validation of AI-Generated Antibody Libraries in 14 Days