SadPhos Library: A Comprehensive Resource for Exploring Chiral Ligand Chemical Space

Traditionally, the discovery of ligands for organic reactions has relied heavily on the intuition and experience of chemists, leading to a trial-and-error process that is both time-consuming and inherently biased. The rise of data science now offers a more systematic and efficient approach to exploring chemical spaces, moving beyond the heuristic constraints of conventional ligand design and enabling a more data-driven, predictive method. In this study, we introduce "SadPhos Library", a comprehensive collection of 890 reported chiral sulfinamide phosphine ligands, and use physical organic descriptors to systematically map their chemical space. By examining a small dataset of known active ligands, we demonstrate how SadPhos library can help identify key properties associated with ligand performance and thus streamline the process of ligand optimization. Furthermore, employing dimensionality reduction and clustering techniques, we pinpoint a representative subset of SadPhos ligands that facilitates more targeted and efficient exploration of this diverse chemical landscape.