A naukri.com initiative
Dev
1w
418
Image Credit: Dev
# SandboxAQ Releases Massive Synthetic Molecule Dataset to Accelerate AI Drug Discovery
- SandboxAQ, an AI and quantum tech startup, released a large-scale dataset of synthetic 3D molecular structures for drug discovery.
- The dataset contains over 5.2 million conformers with diverse chemical scaffolds to aid in training ML models for predicting drug-target interactions.
- It includes 3D atomic coordinates, protein binding affinity, chemical class labels, and structural metadata for geometric deep learning models.
- SandboxAQ collaborated with pharma partners and used quantum and AI simulations to create accurate 3D conformers for common therapeutic targets.
- Developers can integrate the dataset with graph neural networks, transformer-based models, and molecular docking software.
- Modelers can train regression or classification models to predict bioactivity and other properties using the structural information.
- This release signifies a significant synthetic dataset for AI-powered drug discovery to reduce failures and prioritize promising leads.
- The dataset is open for academic and non-commercial research, with potential expansion for industry users via subscription.
- SandboxAQ plans to introduce benchmarks for protein-ligand prediction tasks based on this dataset in the future.
Read Full Article
25 Likes
For uninterrupted reading, download the app