ANI-2x Conformers Recomputed at B97-3c/def2-mTZVP with Forces, ColabFit Conversion
- Provider
- Huddleston, Zubatyuk, Smith, Roitberg, Isayev, Pickering, Devereux, Barros (ANI-2x); ColabFit conversion
- License
- CC-BY-4.0 (source)
- Size
- 98 files
- Format
- Parquet (in sampled files)
- Stored at
- Hugging Face (https://huggingface.co/datasets/colabfit/ANI-2x-B973c-def2mTZVP)
- Access
- Free and public. No account with the data provider is needed.
- DOI
- 10.5281/zenodo.10108942
- Category
- Chemistry & Materials
- Source
- https://doi.org/10.1021/acs.jctc.0c00121
About this dataset
The ANI-2x organic conformer set recomputed with the low-cost composite functional B97-3c and the def2-mTZVP basis in ORCA, one of four levels of theory in the 2023 ANI-2x Zenodo release. Structures have 2 to 63 atoms and contain H, C, N, O, S, F and Cl, drawn from the active-learning sampling of drug-like chemical space that produced ANI-2x. ColabFit counts 9,642,825 configurations and about 146.6 million atoms, labelled with total energies and atomic forces, with force corrections and dipole moments kept in the metadata; configuration sets are split by the number of atoms per structure. Together with the wB97X/6-31G(d), wB97M-V and wB97M-D3(BJ) variants it lets you train or transfer-learn neural network potentials such as ANI, AIMNet2 or MACE across levels of theory on identical geometries. This is the ColabFit Exchange conversion: configurations sit in Parquet files under co/ (positions, cell, species, computed properties and metadata per row) with dataset-level information in ds.parquet; read them with pyarrow, pandas, polars or Hugging Face datasets following the ColabFit schema at materials.colabfit.org/docs. Released by the authors on Zenodo under CC-BY-4.0; cite Devereux et al., J. Chem. Theory Comput. 2020.
How to load this dataset in Python
from datasets import load_dataset
ds = load_dataset("colabfit/ANI-2x-B973c-def2mTZVP")
print(ds)Data structure
File types in sampled files: Parquet
Files
Example file: co/co_0.parquet
Analyze this dataset without downloading it
On Scigantic, ANI-2x Conformers Recomputed at B97-3c/def2-mTZVP with Forces, ColabFit Conversion opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.
This page is also available as Markdown and as JSON.
Related Chemistry & Materials datasets
- Open Molecules 2025 (OMol25), Original ASE-LMDB Release from Meta FAIR Chemistry
- HME21: High-Temperature Multi-Element DFT Training Set for the PFP Universal Potential, ColabFit Conversion
- Defected Phosphorene: Monolayer Black Phosphorus with Vacancies, DFT Data for HDNNP MD, ColabFit Conversion
- Liquid chromatography instrument output with ASM: 14 exports from Agilent OpenLab CDS, Waters Empower (Benchling adapter), Cytiva UNICORN, Thermo Fisher Chromeleon (Benchling adapter)
- GEOM-Drugs: Energy-Annotated Conformer Ensembles for Drug-Like Molecules (SDF)
- QM9 C7H10O2 Constitutional Isomers: 6,095 Molecules at G4MP2 (Ramakrishnan et al. 2014, figshare), ColabFit Conversion
- Alexandria Materials Database Snapshot: 5.07 Million DFT Crystal Entries (pymatgen Parquet)
- Open Catalyst 2025 (OC25) Validation Split: Solid-Liquid Interface DFT, ColabFit Conversion