PAULYLSER Toolbox
Predict highly accurate solute descriptors using our advanced machine learning models, then seamlessly calculate partition coefficients, retention times, and diverse physicochemical properties based on established phase equations.
Empowering Your Predictive Workflows
Our LSER Toolbox bridges the gap between cutting-edge AI and established physical chemistry, giving you versatile and interpretable data.
- Mechanistic Transparency: Combine the predictive power of machine learning with established physical equations. This two-step approach ensures your calculated properties remain highly interpretable and scientifically sound.
- Broad Application Range: Generate versatile solute descriptors that can be applied to diverse chemical systems—from environmental phase partitioning to complex chromatographic separations.
- Intuitive & Flexible Input: Start calculations instantly using standard SMILES strings. Run single predictions or large-scale batches and extract your results effortlessly without the need for rigid CSV templates.
- Streamlined Property Calculation: Once the descriptors are predicted, the software automatically applies the known phase parameters to compute your target properties instantly.
Advanced Modeling for Solute Descriptors
To calculate properties within the Linear Solvation Energy Relationship framework, our toolbox first predicts the critical solute descriptors. You can choose from the following methodologies to generate these descriptors:
- Graph Neural Networks (GNNs): Our best-performing models. These deep learning architectures directly learn from molecular topology to provide the highest accuracy for descriptor prediction across the chemical space.
- Group Contribution Methods: Transparent and well-established baseline models for reliable descriptor estimations.
- k-Nearest Neighbors (k-NN): Fast predictions based on structural similarity.
Once the solute descriptors are predicted using your method of choice, the toolbox utilizes established LSER equations to compute final endpoints like partition coefficients and retention times.
Peer-Review Validated
We don’t just calculate; we validate. Our predictive models for LSER solute descriptors have been rigorously benchmarked and published in peer-reviewed scientific literature.
Publications
- "Prediction of solute descriptors for linear solvation energy relationships using k-nearest neighbours, group contributions, and graph-convolutional neural networks", ACS Environ. Au, 2026. https://doi.org/10.1021/acsenvironau.6c00063