Partial to Total\nGeneration of 3D Transition-Metal\nComplexes

The design of transition-metal\ncomplexes (TMCs) has drawn much\nattention over the years because of their important applications as\nmetallodrugs and functional materials. In this work, we present an\nextension of our recently reported approach, LigandDiff [Jin et al. <i>J. Chem. Theory Comput.</i> <b>20</b>, 4377(2024)]. The\nnew model, which we call multi-LigandDiff, is more flexible and greatly\noutperforms its predecessor. This scaffold-based diffusion model allows\nde novo ligand design with either existing ligands or without any\nligand. Moreover, it allows users to predefine the denticity of the\ngenerated ligand. Our results indicate that multi-LigandDiff can generate\nwell-defined ligands and is transferable to multiple transition metals\nand coordination geometries. In terms of its application, multi-LigandDiff\nsuccessfully designed 338 Fe(II) spin-crossover (SCO) complexes from\nonly 47 experimentally validated SCO complexes. And these generated\ncomplexes are configurationally diverse and structurally reasonable.\nOverall, the results show that multi-LigandDiff is an ideal tool to\ndesign novel TMCs from scratch.