tmQM has become one of the most popular datasets for data-driven research approaches to the chemistry of the transition metals. At present, tmQM is no longer an isolated dataset but a series of datasets providing geometries, graphs, ligand libraries, electron excitation data, and multi-level quantum properties. As this ecosystem keeps growing, the integration of this data for its use in machine learning and generative AI models becomes more and more difficult. Responding to this challenge, Luca Cibinel developed a knowledge graph representation that uses a resource description framework vocabulary to express detailed descriptions of transition metal complexes. These descriptions are both qualitative and quantitative in their nature, and they are sourced from the different members of the tmQM dataset family. The research was based on an interdisciplinary collaboration between the SFF Centers INTEGREAT (Luca Cibinel, Johan Pensar, Riccardo De Bin and Basil Ell) and Hylleraas (David Balcells), and the ICReDD Institute (Hylleraas alumnus Trond Linjordet).
Reference
Cibinel, L.; Linjordet, T.; Pensar, J.; Balcells, D.; De Bin, R.; Ell, B. tmQM-RDF Data Set: A Knowledge Graph Representing Transition Metal Complexes J. Chem. Inf. Model. 2026, 66, 7524.
Â