This set of catalysts includes all those catalysts that were synthesised and tested experimentally by our labs. The figures published in the paper can be reproduced using the Jupyter notebook Analysis_MS_figures.ipynb. A more general-purpose analysis of the AF scores can be performed with Analysis_test_set_1.ipynb.
The files are named according to the following internal convention, which is encoded in the scripts provided under ../../src/. Here, name is a string that identifies the input to the catalyst evaluator.
name.sdfcontains the graph-based definition of the ligand set; This is the very input to the catalyst evaluator.name_out.sdfis the final output of the catalyst evaluator; it contains the values of the fitness score, descriptors, and weights.name_outSubPreDFT-D.sdfcontains the geometry of the metallacyclobutane intermediate (DFT-based optimization).name_outSubPreDFT-X.sdfcontains the transition state model for productive metathesis reaction (DFT-based constrained optimization).name_outSubPreDFT-Z.sdfcontains the transition state model for beta-H elimination reaction (DFT-based constrained optimization).name_outSubXTB-A.sdfcontains the geometry of the catalyst precursor (xTB-based optimization).name_outSubXTB-C.sdfcontains the geometry of the metallacyclobutane intermediate (xTB-based optimization).name_outSubXTB-E.sdf(Optional: produced only when X ligands differ from Cl) contains the geometry of the catalyst precursor with trans-X configuration and X=Cl (xTB-based optimization).name_outSubXTB-F.sdfcontains the geometry of the catalyst precursor with cis-X configuration (xTB-based optimization).name_outSubXTB-L.sdfcontains the geometry of the free dative ligand (xTB-based optimization).