Around Around
Sign in

🧬 REINVENT4 Molecular Design

AI-Driven Molecule Generation Platform v4.7.0

De Novo Generation
  • Generate completely new molecules from scratch
  • Uses Reinvent prior model (trained on ChEMBL 25)
  • Supports Sampling (UI) — RL handled later
Select Generation Mode
Basic Parameters
Higher = more diverse
🎮 Reinforcement Learning Settings
Molecule Optimization
  • Generate optimized analogs of an existing molecule
  • Uses Mol2Mol similarity prior
  • UI sends the first SMILES line (backend expects single string)
Select Generation Mode
Input Molecule
Backend endpoint accepts a single SMILES string; UI sends first line.
🎮 Reinforcement Learning Settings
Scaffold Hopping
  • Replace molecular scaffolds while preserving R-groups
  • Uses Libinvent prior
  • Use [*] for attachment points
Select Generation Mode
Input Scaffold
Use [*] to mark R-group attachment points.
🎮 Reinforcement Learning Settings
Generating molecules...

This may take several minutes depending on the number of molecules and mode selected.

Generated Molecules