QM9x: QM9 Recomputed at wB97X/6-31G(d) with Forces, ColabFit Conversion

Provider
Mathias Schreiner, Arghya Bhowmik, Tejs Vegge, Jonas Busk, Ole Winther (DTU); ColabFit conversion
License
CC-BY-4.0 (source)
Size
4 files
Format
Parquet (in sampled files)
Stored at
Hugging Face (https://huggingface.co/datasets/colabfit/QM9x)
Access
Free and public. No account with the data provider is needed.
DOI
10.60732/1edbb6e0
Category
Chemistry & Materials
Source
https://doi.org/10.1038/s41597-022-01870-w

About QM9x: QM9 Recomputed at wB97X/6-31G(d) with Forces, ColabFit Conversion

QM9x recomputes energies and forces for the QM9 molecules at the wB97X/6-31G(d) level of theory, the same level used for Transition1x and ANI-1x, so that QM9 can be compared directly with reaction-path data. Because the recomputation shifts the potential energy surface, most QM9 equilibrium geometries carry non-zero forces at this level, which makes the set usable for training and testing machine learning interatomic potentials rather than only property regression. Published by Schreiner, Bhowmik, Vegge, Busk and Winther (DTU) alongside Transition1x in Scientific Data 2022. ColabFit's card counts 133,871 configurations (C, H, N, O, F) with energy, atomization energy and atomic forces; the repository is about 0.4 GB (measured 2026-11-14). This is the ColabFit Exchange conversion: configurations are Parquet files under co/ (structure, positions, cell, 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 configuration schema at materials.colabfit.org/docs. The upstream figshare record is CC-BY-4.0.

How to load this dataset in Python

from datasets import load_dataset
ds = load_dataset("colabfit/QM9x")
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, QM9x: QM9 Recomputed at wB97X/6-31G(d) 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