yanliang3612/NucleusDiff
[PNAS 2025] Code of "Manifold-Constrained Nucleus-Level Denoising Diffusion Model for Structure-Based Drug Design".
This project helps drug designers accelerate the discovery of new drug candidates by generating novel small molecules that precisely fit into specific protein binding sites. It takes the 3D structure of a protein's binding pocket as input and outputs potential new drug-like molecules that are structurally optimized to bind effectively. Medicinal chemists and computational biologists focused on lead generation and optimization in drug discovery would use this.
Use this if you need to computationally design new small molecules that accurately dock into a target protein's binding site for drug discovery.
Not ideal if you are looking for a simple, off-the-shelf software with a graphical user interface, as this requires a technical setup and scripting.
Stars
32
Forks
3
Language
Python
License
MIT
Category
Last pushed
Jan 18, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/yanliang3612/NucleusDiff"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.