Apheris logo
Menu

Case Study: AI Structural Biology (AISB) Network fine-tunes OpenFold3 (Columbia University) on Proprietary Pharma Data

In the past two years, foundational AI models for structural biology have made important technical progress. Despite these advances, current models still show limited accuracy in areas that are central to drug discovery. The AI Structural Biology (AISB) Network - a collective of large biopharmaceutical companies and Apheris as a federation technology provider - was formed to address these barriers.


In this case study we explore:


  • The challenges around access to high-quality molecular data for Life Sciences
  • How the Network consisting of major pharmaceutical companies, prestigious model building partners and Apheris as the technology provider was created to tackle this
  • Dive into an example initiative: Fine-tuning OpenFold3 on public pharma data to improve accuracy and generalizability

Recommended for you

Privacy-preserving Data Ecosystems in Support of Drug Discovery
In recent years, Deep Learning has gained considerable traction in many fields, but none more so than in the realm of drug discovery. Applications range from generating de novo hit-like molecules, predicting drug-disease associations and activity, toxicology estimation, to the analysis of medical images.

In this white paper, you will learn:

checkmark How you can preserve intellectual property while generating value from sensitive data
checkmark Deep dive into federated and secure QSAR models
checkmark How Apheris prevents uncontrolled data sharing
checkmark How you can pave the way to precision medicine with federated and secure QSAR model training
Privacy-preserving Data Science on QSAR models
This case study shows the effectiveness of Apheris’ platform and services for data science on data that is not directly accessible and distributed. Federated learning decentralizes learning algorithms to access multiple data sets, maintaining privacy. Without pooling sensitive data, new insights can be uncovered securely.

In this white paper, you will learn:

checkmark How federated learning, privacy testing and other tools that enable data science on not directly accessible and distributed data
checkmark Sowcase of federated learning, privacy testing and other tools that enable data science practices on not directly accessible and distributed data, at the example of QSAR models.
Computational Governance & healthcare data: pioneering new frontiers within regulatory boundaries
Computational governance enables machine learning on European healthcare data.
European hospitals face a unique challenge, balancing complex regulatory frameworks with the need to make valuable data available for research. Computational governance exists as a solution and enables compliant research on health data.