Abstract
In this work, we introduced a siderophore information database (SIDERTE), a digitized siderophore information database containing 649 unique structures. Leveraging this digitalized data set, we gained a systematic overview of siderophores by their clustering patterns in the chemical space. Building upon this, we developed a functional group-based method for predicting new iron-binding molecules with experimental validation. Expanding our approach to the collection of open natural products (COCONUT) database, we predicted a staggering 3199 siderophore candidates, showcasing remarkable structure diversity that is largely unexplored. Our study provides a valuable resource for accelerating the discovery of novel iron-binding molecules and advancing our understanding of siderophores.