We are glad to announce a groundbreaking publication in the Journal of Chemical Theory and Computation.

This research addresses a critical challenge in computational protein engineering: the scarcity of high-quality protein data. By leveraging transfer learning, Xiangwen and Jiahui developed a framework that incorporates knowledge from larger datasets to enhance predictions on smaller, specialized datasets. Their approach combines machine learning with structural features from enzymes and substrates, enabling accurate modeling of enzyme–substrate interactions. The use of transfer learning ensures that the model generalizes well, even when training data is limited—a common hurdle in this field.

The full article can be accessed here: https://pubs.acs.org/doi/full/10.1021/acs.jctc.4c01391.