A Nice Smile: Predicting Chemical Properties from SMILES Notation (2025-JA-25MP-2678)In recent years, quantitative structure–activity relationship (QSAR) techniques have been developed to build advanced models using machine learning to predict the properties of new molecules. This approach is expected to reduce experimental time and costs. In this presentation, we will introduce the Materials Informatics Toolkit and Torch Deep Learning extensions from the JMP Marketplace. These add-ins utilize the widely used open-source RDKit package to process SMILES strings of chemical substances, providing chemists with an interactive and powerful workflow for performing QSAR. This presentation will show some examples of using these extensions.
Automation and Scripting
Predictive Modeling and Machine Learning
