Closing the Loop with Amazon Bio Discovery’s Integrated Lab Partners
May 7, 2026
Read the blog
Introducing Amazon Bio Discovery: AI-Powered Drug Discovery at Scale
April 14, 2026
Product Infographic: The DBTL Workflow
April 14, 2026
AWS and Gray Lab at Johns Hopkins announce groundbreaking database for AI/ML antibody design
Available Foundation Models
Bring Your Own Model (BYOM)
Bring any models your organization has developed in-house or licensed from commercial model providers and integrate them seamlessly into Amazon Bio Discovery workflows. Run your proprietary models alongside our catalog of 40+ biological AI models in unified pipelines, enabling you to leverage your existing IP while benefiting from Amazon Bio Discovery's lab-in-the-loop infrastructure. This feature is currently in beta.
Train Your Own Model
Fine-tune biological AI models with your existing experimental data or newly generated results to improve predictions with every iteration. Upload proprietary datasets and train on your experimental results while maintaining full data ownership — your data is never used for hosted model training, ensuring complete control over your intellectual property. This feature is currently in beta.
Agent-Guided De Novo Design of Nanobody Binders Against a Novel Cancer Target
Date: April 2026
Abstract: Therapeutic antibody discovery remains slow and resource-intensive, with traditional methods providing limited control over epitope selection. We present a workflow for de novo nanobody design applied to a novel Desmoplastic Small Round Cell Tumor target encompassing four stages: epitope identification, de novo generation, multi-metric scoring, and high-throughput screening. We generated 288,000 nanobody designs spanning eight target epitope regions. Multi-objective Pareto filtering yielded 100,000 candidates for YSD screening. 46/116 candidates (39.7%) produced reliable kinetic fits with KD values from 0.66 nM to 305 nM (median 31.7 nM). These results show that an agent-guided computational workflow can design nanomolar to sub-nanomolar nanobody binders against a novel target without experimental structure or prior antibody information.
Context-Aware Multi-Property Antibody Predictor
Date: January 2026
Abstract: Recent advances in Machine Learning have transformed antibody development through in-silico models. We introduce a novel multi-modal architecture featuring specialized tokenization and embedding projection that integrates text and protein language models (pLM) with a learning strategy to enable in-context learning for multi-property prediction. Using 876,898 antibodies with batch effect simulation, our architecture achieved Spearman's ρ>0.8 across multiple developability properties.
Getting Started with Amazon Bio Discovery
Get up to speed quickly with structured learning designed for bench scientists, computational biologists, and structural biologists. This comprehensive course covers product overview, running your first experiment, model selection, workflow building, and best practices for drug discovery workflows.
