1,497 unique AI-designed GID4 (CTLH E3 ligase / TPD) binders as docked protein–ligand complexes

#3
by Tc-43 - opened
Hugging Science org

Hi all — sharing an open-source generative chemistry dataset that might interest the structure/drug-discovery folks here: 1,497 unique AI-designed GID4 (CTLH E3 ligase / TPD) binders as docked protein–ligand complexes, CC-BY-4.0.

Each entry is a self-contained PDB (receptor + 3D docked pose + 2D↔3D atom map), and there's a loadable Parquet table (SMILES + RDKit descriptors) so you can preview it in the Dataset Viewer or pull it in one line:
load_dataset("Tc-43/GID4_7SLZ_20260620")

📦 Dataset: https://huggingface.co/datasets/Tc-43/GID4_7SLZ_20260620
🔭 3D Space demo: https://huggingface.co/spaces/Tc-43/gid4-7slz-explorer
🗂️ Full collection (9 targets — oncology, immunology, TPD): https://huggingface.co/collections/Tc-43/technetium-ga-ii-generative-design-sets-6a3c0385ed11bfca1fdcfb07
These are computational designs/poses (not experimentally validated — no activity claims), generated by our Technetium GA-II platform. Would love feedback, and happy to contribute it to a Hugging Science collection if useful. 🙏

Sign up or log in to comment