Dizyme: The Ultimate Resource for Nanozyme Multiple Catalytic Activity Prediction

The study introduces DiZyme, a web platform utilizing machine learning to predict multiple catalytic activities of nanozymes, which are materials with significant potential in various fields like biomedicine and environmental monitoring. The platform features a comprehensive database of 1,210 experimental nanozyme samples and employs an ensemble learning algorithm that achieves R² values of 0.75 and 0.77 for predicting the Michaelis-Menten constant and maximum velocity, respectively. DiZyme also includes an assistant powered by a large language model to provide users with detailed information on experimental procedures. This tool aims to facilitate the discovery of new nanozymes and their applications by accurately predicting activities such as peroxidase, oxidase, and catalase.