Datasets:
proteinglm
commited on
Commit
•
faf9d67
1
Parent(s):
410edca
Update README.md
Browse files
README.md
CHANGED
@@ -26,4 +26,74 @@ configs:
|
|
26 |
path: data/valid-*
|
27 |
- split: test
|
28 |
path: data/test-*
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
29 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
26 |
path: data/valid-*
|
27 |
- split: test
|
28 |
path: data/test-*
|
29 |
+
license: apache-2.0
|
30 |
+
task_categories:
|
31 |
+
- text-classification
|
32 |
+
tags:
|
33 |
+
- biology
|
34 |
+
- medical
|
35 |
+
- chemistry
|
36 |
+
size_categories:
|
37 |
+
- 10K<n<100K
|
38 |
---
|
39 |
+
|
40 |
+
|
41 |
+
# Dataset Card for Fold Prediction Dataset
|
42 |
+
|
43 |
+
### Dataset Summary
|
44 |
+
|
45 |
+
Fold class prediction is a scientific classification task that assigns protein sequences to one of 1,195 known folds. The primary application of this task lies in the identification of novel remote homologs among proteins of interest, such as emerging antibiotic-resistant genes and industrial enzymes. The study of protein fold holds great significance in fields like proteomics and structural biology, as it facilitates the analysis of folding patterns, leading to the discovery of remote homologies and advancements in disease research.
|
46 |
+
|
47 |
+
## Dataset Structure
|
48 |
+
|
49 |
+
### Data Instances
|
50 |
+
For each instance, there is a string representing the protein sequence and an integer label indicating which know fold a protein sequence belongs to. See the [fold prediction dataset viewer](https://huggingface.co/datasets/Bo1015/fold_prediction/viewer) to explore more examples.
|
51 |
+
|
52 |
+
```
|
53 |
+
{'seq':'MEHVIDNFDNIDKCLKCGKPIKVVKLKYIKKKIENIPNSHLINFKYCSKCKRENVIENL'
|
54 |
+
'label':6}
|
55 |
+
```
|
56 |
+
|
57 |
+
The average for the `seq` and the `label` are provided below:
|
58 |
+
|
59 |
+
| Feature | Mean Count |
|
60 |
+
| ---------- | ---------------- |
|
61 |
+
| seq | 168 |
|
62 |
+
|
63 |
+
|
64 |
+
|
65 |
+
|
66 |
+
### Data Fields
|
67 |
+
|
68 |
+
- `seq`: a string containing the protein sequence.
|
69 |
+
- `label`: an integer label indicating which know fold a protein sequence belongs to.
|
70 |
+
|
71 |
+
### Data Splits
|
72 |
+
|
73 |
+
The fold prediction dataset has 3 splits: _train_, _valid_ and _test_. Below are the statistics of the dataset.
|
74 |
+
|
75 |
+
| Dataset Split | Number of Instances in Split |
|
76 |
+
| ------------- | ------------------------------------------- |
|
77 |
+
| Train | 12,312 |
|
78 |
+
|Valid | 736|
|
79 |
+
| Test | 3,244 |
|
80 |
+
|
81 |
+
### Source Data
|
82 |
+
|
83 |
+
#### Initial Data Collection and Normalization
|
84 |
+
The dataset employed for this task is based on [SCOP 1.75](https://scop.mrc-lmb.cam.ac.uk/), a release from 2009.
|
85 |
+
|
86 |
+
### Citation
|
87 |
+
If you find our work useful, please consider citing the following paper:
|
88 |
+
|
89 |
+
```
|
90 |
+
@misc{chen2024xtrimopglm,
|
91 |
+
title={xTrimoPGLM: unified 100B-scale pre-trained transformer for deciphering the language of protein},
|
92 |
+
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},
|
93 |
+
year={2024},
|
94 |
+
eprint={2401.06199},
|
95 |
+
archivePrefix={arXiv},
|
96 |
+
primaryClass={cs.CL},
|
97 |
+
note={arXiv preprint arXiv:2401.06199}
|
98 |
+
}
|
99 |
+
```
|