Graphic titled A Hybrid AI Model for Resolving Molecular Structures Integrating Sequence and Cryo-Electron Microscopy Data, with the PUXANO logo and protein structures

PUXANO Leads Innovation Study on Hybrid AI Model for Molecular Structure Resolution

27 August 2025

PUXANO is leading an innovative study as part of the FFplus program, titled “Hybrid AI-Model for Resolving Molecular Structures Integrating Sequence and Cryo-Electron Microscopy Data,” driving forward the integration of AI and structural biology.

This study focuses on developing and training a hybrid de novo protein structure prediction model, built on the Boltz/OpenFold3 framework. By integrating cryo-EM data with protein sequence information, the model aims to reduce structure hallucinations often observed in AI systems like AlphaFold3. Furthermore, the model will accelerate epitope mapping projects by screening 12-24 mAbs in parallel in a cost-efficient manner.

This project is driven in collaboration with:

  • CSIC: contributing cutting-edge expertise in image processing

  • IT4Innovations National Supercomputing Center: enabling high-performance computing

Read more about this project.

 
Interested in learning more?

Fill out this form or contact us via info@puxano.com.

Read also

Sign Up for Our Newsletter

Stay updated on the latest in protein design, AI-driven engineering, and groundbreaking research. Subscribe now to get exclusive insights and updates.