qyp111/mdCATH

---
viewer: false
license: cc-by-4.0
tags:
- chemistry
- biology
- molecular dynamics
- neural network potential
pretty_name: 'mdCATH: A Large-Scale MD Dataset for Data-Driven Computational Biophysics'
author: A. Mirarchi, T. Giorgino and G. De Fabritiis
size_categories:
- 10M<n<100M
---

# mdCATH: A Large-Scale MD Dataset for Data-Driven Computational Biophysics

This dataset comprises all-atom systems for 5,398 CATH domains, modeled with a state-of-the-art classical force field, and simulated in five replicates each at five temperatures from 320 K to 450 K.

## Availability
- [torchmd-net dataloader](https://github.com/torchmd/torchmd-net/blob/main/torchmdnet/datasets/mdcath.py)
- [playmolecule](https://open.playmolecule.org/mdcath)
- [scripts to load, convert and rebuild](https://github.com/compsciencelab/mdCATH)

## Citing The Dataset
Please cite this manuscript for papers that use the mdCATH dataset:
> Mirarchi, A., Giorgino, T. & De Fabritiis, G. mdCATH: A Large-Scale MD Dataset for Data-Driven Computational Biophysics. Sci Data 11, 1299 (2024). https://doi.org/10.1038/s41597-024-04140-z. Preprint available at [arXiv:2407.14794](https://arxiv.org/abs/2407.14794v1) (2024).

## Dataset Size

| Description | Value |
|:---------------------|:-------------|
| Domains | 5,398 |
| Trajectories | 134,950 |
| Total sampled time | 62.6 ms |
| Total atoms | 11,671,592 |
| Total amino acids | 740,813 |
| Avg. traj. length | 464 ns |
| Avg. system size | 2,162 atoms |
| Avg. domain length | 137 AAs |
| Total file size | 3.3 TB |