AutoMulti: An Integrated Graphical Platform for Computational Chemistry Analysis and Molecular Data Generation.

Computational chemistry supports molecular interpretation and the generation of reusable data sets for statistical modeling and machine learning. However, molecular construction, quantum-chemical input preparation, result analysis, reference-data retrieval, and descriptor generation are often handled by separate programs and assembled manually. AutoMulti is a graphical platform that connects these operations in one desktop environment. In a diimine-Ni catalyst case study, AutoMulti was used for catalyst enumeration, batch GFN2-xTB calculations, QM-feature and descriptor extraction, and structured data export. The exported data set was subsequently analyzed in an independent graph neural network (GNN) workflow, which reproduced the GFN2-xTB-derived IM/TS energy-difference labels.