SciGroveBeta
Cheminformatics

JEDEL: Zero-Shot DNA-Encoded Library Design for Early-Stage Drug Discovery

Zygimantas Jocys, Zhanxing Zhu, Henriette M.G. Willems, Katayoun Farrahi

Featured June 29, 2026

Original Paper

AI-generated analysis — This is SciGrove's AI interpretation of the paper, not peer-reviewed content. Always refer to the original paper.

Simply

This new method helps scientists design new medicines by creating synthesizable drug recipes directly from a target's desired interaction pattern, making drug discovery faster and more efficient.

In depth
The paper introduces JEDEL, a novel framework that directly translates 3D pharmacophore representations of active ligands into synthesis-ready DNA-encoded libraries (DELs). Unlike prior generative models that produce virtual compounds, JEDEL ensures every output is experimentally realizable by operating within the space of purchasable building blocks and validated reactions, enabling pharmacophore-guided combinatorial library design at scale.

Key Takeaways

  • 1
    JEDEL is the first framework to generate synthesis-ready DELs directly from 3D pharmacophore representations, guaranteeing experimental realizability by design.
  • 2
    The model employs a Joint Embedding Predictive Architecture (JEPA) and a hierarchical decoder to efficiently map pharmacophore geometry to molecular structure and scalable synthesis instructions.
  • 3
    JEDEL demonstrates zero-shot capability across 18 diverse protein targets, outperforming baselines in predicted binding affinity, pharmacophore recovery, and sample efficiency without target-specific retraining.

Conceptual Flow

HIGH LEVEL
1
Methodology: Designing Synthesizable Drugs

The system takes a desired drug shape, figures out what building blocks are needed, and then gives instructions to build the drug.

Desired Drug Shape
Figure Out Parts
Building Instructions
2
Results: More Effective Drug Candidates

Compared to old ways, this new method finds more good drug candidates using fewer tries, saving time and resources.

Old Way: Many Tries
New Way: Few Tries
Find Good Drugs
Old Way: Few Good Drugs
New Way: Many Good Drugs

This breakdown was generated by SciGrove. Get the same analysis — intuition, storyboard, peer review, a runnable prototype and a glossary — on any paper you upload or paste a DOI for.