Project title
OmniFold: Open-Source Biomolecular Structure Prediction for All
Collaborators and funding
-
Alex Morehead and Jianlin Cheng, University of Missouri
-
Milot Mirdita and Martin Steinegger, Seoul National University.
-
Hoa La and Minjia Zhang, Department of Computer Science, University of Illinois Urbana-Champagne
-
George Bouras, Adelaide Medical School, Faculty of Health and Medical Sciences, University of Adelaide
-
Robert Edwards Flinders University. NCMAS 2025-14 (awarded to Robert Edwards CIA)
Contact(s)
-
Alex Morehead, University of Missouri, alex.morehead@gmail.com
-
George Bouras, Adelaide Medical School, Faculty of Health and Medical Sciences, University of Adelaide, george.bouras@adelaide.edu.au
Project description and aims
This project aims to develop an open-source replication of DeepMind’s Alphafold3 (AF3), the state of the art tool used for biomolecular structure prediction following DeepMind’s breakthrough protein structure prediction tool AlphaFold2, for which the Nobel Prize for Chemistry was awarded in 2024. Along with replicating AF3 in an open-source and user friendly manner, by making changes to the underlying architecture of the model, we hope to surpass its performance on certain problems and molecule types.
How is ABLeS supporting this work?
This work is supported through the software accelerator scheme provided by ABLeS. The supports includes 100 KSUs per quarter.
Expected outputs enabled by participation in ABLeS
Model weights and full inference pipeline will be available via GitHub and HuggingFace.
These details have been provided by project members at project initiation. For more information on the project, please consult the contact(s) or project links above.