Sign in to access this content
Model
MolMIM allows users to generate molecules similar to the seed molecule in SMILES format by randomly perturbing the latent space encoded from a seed molecule and decoding that back into SMILES.
Sign in to access this content
| Field | Response |
|---|---|
| Intended Application(s) & Domain(s): | Molecular drug discovery and design |
| Model Type: | Molecular Sequence Generation |
| Intended Users: | This model is intended for developers in the academic or pharmaceutical industries who build artificial intelligence applications to perform property guided molecule optimization and novel molecule generation. |
| Outputs: | Text (Molecule Sequence) |
| Describe how the model works: | Computes numerical embeddings for molecular representations and generates similar molecular representations from numerical embeddings |
| Technical Limitations: | Model may not perform well on sequences that are highly divergent from the ZINC-15 dataset. |
| Verified to have met prescribed NVIDIA standards: | Yes |
| Performance Metrics: | * Similarity * Modelability * Nearest Neighbor Correlation * Validity * Unique * Novelty * Non Identicality * Effective Novelty * Scaffold Unique * Scaffold Non-Identical Similarity * Scaffold Novelty * Effective Scaffold Novelty * Entropy |
| Potential Known Risks: | The model may produce molecules that are invalid; validate with RDKit. |
| Licensing: | https://docs.nvidia.com/ai-foundation-models-community-license.pdf |