Data-Driven Formulation Based on Integrated Symbolic Regression of Hydrogel Swelling and Molecular Interactions

Hydrogels offer promising solutions across various fields. However, understanding complex behaviors like swelling and ligand interaction requires multiperspective data, from functional groups to molecular dynamics. Single-perspective analyses often fall short, especially when ligand-induced selective adsorption occurs. This study presents a data-driven approach integrating TD-NMR, 1H–15N HSQC, 1H–13C HNCO with isotope-labeled peptides, RDKit descriptors, and DSC data. Using symbolic regression, we derived highly accurate, interpretable equations (e.g., swelling ratio accuracy = 1.0 on test set). This methodology reveals fundamental hydrogel mechanisms and provides a framework for rational design.