QC Fitting Datasets for OpenFF SMIRNOFF Sage 2.3.0

Description A quantum chemical (QC) optimization and torsiondrive datasets generated at the OpenFF default level of theory, B3LYP-D3BJ/DZVP, and curated to train parameters in OpenFF 2.3.0 Sage with NAGL partial charge model AshGC v1.0. The records compiled in this dataset are a subset sourced from the QCArchive datasets listed in the Technical Information section, were there is overlap with the datasets used to train previous versions of Sage (see references). Additional details can be found in the GitHub dataset repository and the Force field repository. General Information * Date: 2026-01-27 * Name: OpenFF SMIRNOFF Sage 2.3.0 * Dataset submitter: Jennifer A Clark * Dataset curator: Lily Wang * Class: OpenFF Optimization Dataset * Dataset Type: optimization * Purpose: B3LYP-D3BJ/DZVP conformers for training OpenFF 2.3.0 Sage with AshGC v1.0 NAGL partial charge model. * Number of unique molecules: 4696 * Number of conformers: 4696 * Number of conformers (min, mean, max): 1.00, 1.00, 1.00 * Molecular weight (min, mean, max): 32.05, 207.67, 878.25 * Charges: -4.0, -3.0, -2.0, -1.0, 0.0, 1.0, 2.0 * Class: OpenFF TorsionDrive Dataset * Dataset Type: torsiondrive * Purpose: B3LYP-D3BJ/DZVP conformers for training torsion drives for OpenFF 2.3.0 Sage with AshGC v1.0 NAGL partial charge model. * Number of unique molecules: 1265 * Number of driven torsions: 1371 * Number of conformers: 1265 * Number of conformers (min, mean, max): 1, 1, 1 * Molecular weight (min, mean, max): 32.05, 165.33, 511.27 * Charges: -3.0, -2.0, -1.0, 0.0, 1.0, 2.0