Open3dqsar

Run with:

Add exporter to MoleculeDataset pipeline so MIF can be included with other 3D descriptors.
Enable use as input features for scikit-learn, XGBoost, PyTorch models.
Allow feature importance mapping back onto the 3D grid for interpretability (visualize with py3Dmol or open3D).
Provide CLI flag (e.g., --descriptor mif) in existing feature extraction scripts.

Drug Design: Open3DQSAR can be used to design new drugs with specific biological activities. By developing 3DQSAR models that capture the relationships between molecular structure and biological activity, researchers can identify novel lead compounds with improved potency and selectivity.
Optimization of Existing Leads: The software can also be used to optimize existing lead compounds, by identifying structural modifications that improve their biological activity.
Toxicity Prediction: Open3DQSAR can be used to predict the toxicity of molecules, which is essential for ensuring the safety of new drugs.
Material Science: The software has applications in material science, where it can be used to design new materials with specific properties.