# Solvated Protein Fragments (PhysNet Benchmark, revPBE-D3), ColabFit Conversion

## Facts

- Provider: Unke and Meuwly, University of Basel; converted by ColabFit Exchange
- License: CC-BY-4.0 (https://zenodo.org/api/records/2605372)
- Size: 6 files
- Format: Parquet (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/colabfit/solvated_protein_fragments_JCTC_2019
- Access: public, readable without an account
- DOI: https://doi.org/10.60732/c4731f07
- Category: Chemistry & Materials
- Source: https://doi.org/10.1021/acs.jctc.9b00181
- Page: https://scigantic.com/archive/d11c685b-4edd-5431-8f53-2bd5300f3a09
- Keywords: Chemistry & Materials, benchmark, Parquet, Hugging Face

## About

ColabFit Exchange conversion of the solvated protein fragments dataset, published by Oliver T. Unke and Markus Meuwly at the University of Basel (Journal of Chemical Theory and Computation 2019) alongside the SN2 reactions set as a benchmark for PhysNet and other machine learning models of chemical reactions, long-range interactions and condensed phase systems. It contains all possible amons (hydrogen-saturated covalently bonded fragments) of up to eight heavy atoms (C, N, O, S) derivable from proteins of the 20 natural amino acids joined by peptide bonds or disulfide bridges, including (de)protonated charge states up to a total charge of +/-2e. Upstream reference energies, forces and dipole moments for 2,731,180 structures were computed at the revPBE-D3(BJ)/def2-TZVP level of theory with ORCA 4.0.1. ColabFit lists 2,730,942 configurations with energy and atomic forces. Measured 2026-11-13: ds.parquet plus a co/ folder of 3 parquet files (2.2 GB); repository storage 4.8 GB. Rows follow the ColabFit configuration schema (positions, atomic_numbers, cell, pbc, energy, atomic_forces, cauchy_stress, software, method and per-row metadata) and open with pandas.read_parquet, pyarrow or datasets.load_dataset; ds.parquet holds the dataset summary. Licensed CC BY 4.0; cite Unke and Meuwly 2019 and the ColabFit Exchange.

## Load the data

```python
from datasets import load_dataset
ds = load_dataset("colabfit/solvated_protein_fragments_JCTC_2019")
print(ds)
```

## Data structure

File types in sampled files: Parquet

Example file: `co/co_0.parquet`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/d11c685b-4edd-5431-8f53-2bd5300f3a09
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/d11c685b-4edd-5431-8f53-2bd5300f3a09
- Over MCP: `get_schema_card` with id `d11c685b-4edd-5431-8f53-2bd5300f3a09` at https://api.scigantic.com/mcp

## Related topics

- [Parquet datasets](https://scigantic.com/topics/parquet)
- [Benchmark datasets](https://scigantic.com/topics/benchmark)
- [Chemistry & Materials datasets](https://scigantic.com/topics/chemistry-and-materials)
