---
dataset_info:
  features:
  - name: seq
    dtype: string
  - name: label
    dtype: float64
  splits:
  - name: train
    num_bytes: 3068946
    num_examples: 53614
  - name: valid
    num_bytes: 155744
    num_examples: 2512
  - name: test
    num_bytes: 709292
    num_examples: 12851
  download_size: 2058102
  dataset_size: 3933982
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: valid
    path: data/valid-*
  - split: test
    path: data/test-*
license: apache-2.0
task_categories:
- token-classification
tags:
- chemistry
- biology
size_categories:
- 10K<n<100K
---


# Dataset Card for Stability Stability Dataset

### Dataset Summary

The Stability Stability task is to predict the concentration of protease at which a protein can retain its folded state. Protease, being integral to numerous biological processes, bears significant relevance and a profound comprehension of protein stability during protease interaction can offer immense value, especially in the creation of novel therapeutics.

## Dataset Structure

### Data Instances
For each instance, there is a string representing the protein sequence and a float number indicating the stability score. See the [stability prediction dataset viewer](https://huggingface.co/datasets/Bo1015/stability_prediction/viewer) to explore more examples.

```
{'seq':'MEHVIDNFDNIDKCLKCGKPIKVVKLKYIKKKIENIPNSHLINFKYCSKCKRENVIENL'
'label':0.17}
```

The average  for the `seq` and the `label` are provided below:

| Feature    | Mean Count |
| ---------- | ---------------- |
| seq    |    45   |
| label    |    0.34   |



### Data Fields

- `seq`: a string containing the protein sequence
- `label`: a float number indicating the stability score of each sequence.

### Data Splits

The stability stability dataset has 3 splits: _train_, _valid_, and _test_. Below are the statistics of the dataset.

| Dataset Split | Number of Instances in Split                |
| ------------- | ------------------------------------------- |
| Train         | 53,614                          |
| Valid         | 2,512                              |
| Test          | 12,851                                      |

### Source Data

#### Initial Data Collection and Normalization

The dataset applied in this task is initially sourced from [Rocklin et al](https://www.science.org/doi/10.1126/science.aan0693) and subsequently collected within the [TAPE](https://github.com/songlab-cal/tape).

### Licensing Information

The dataset is released under the [Apache-2.0 License](http://www.apache.org/licenses/LICENSE-2.0). 

### Citation
If you find our work useful, please consider citing the following paper:

```
@misc{chen2024xtrimopglm,
  title={xTrimoPGLM: unified 100B-scale pre-trained transformer for deciphering the language of protein},
  author={Chen, Bo and Cheng, Xingyi and Li, Pan and Geng, Yangli-ao and Gong, Jing and Li, Shen and Bei, Zhilei and Tan, Xu and Wang, Boyan and Zeng, Xin and others},
  year={2024},
  eprint={2401.06199},
  archivePrefix={arXiv},
  primaryClass={cs.CL},
  note={arXiv preprint arXiv:2401.06199}
}
```