Bo1015 commited on
Commit
fd60679
·
verified ·
1 Parent(s): 0a33d79

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +65 -0
README.md CHANGED
@@ -37,4 +37,69 @@ size_categories:
37
  - 1K<n<10K
38
  ---
39
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
40
 
 
37
  - 1K<n<10K
38
  ---
39
 
40
+ # Dataset Card for Contact Prediction Dataset
41
+
42
+ ### Dataset Summary
43
+
44
+ Contact map prediction aims to determine whether two residues, $i$ and $j$, are in contact or not, based on their distance with a certain threshold ($<$8 Angstrom). This task is an important part of the early Alphafold version for structural prediction.
45
+
46
+ ## Dataset Structure
47
+
48
+ ### Data Instances
49
+
50
+ For each instance, there is a string of the protein sequences, a sequence for the contact labels. The each of the sub-labels "[2, 3]" indicates the 3rd residue are in contact with the 4th residue (start from index 0). See the [Contact map prediction dataset viewer](https://huggingface.co/datasets/Bo1015/contact_prediction_binary/viewer/default/test) to explore more examples.
51
+
52
+ ```
53
+ {'seq':'QNLLKNLAASLGRKPFVADKQGVYRLTIDKHLVMLAPHGSELVLRTPIDAPMLREGNNVNVTLLRSLMQQALAWAKRYPQTLVLDDCGQLVLEARLRLQELDTHGLQEVINKQLALLEHLIPQLTP'
54
+ 'label': [ [ 0, 0 ], [ 0, 1 ], [ 1, 1 ], [ 1, 2 ], [ 1, 3 ], [ 1, 101 ], [ 2, 2 ], [ 2, 3 ], [ 2, 4 ], [ 3, 3 ], [ 3, 4 ], [ 3, 5 ], [ 3, 99 ], [ 3, 100 ], [ 3, 101 ], [ 4, 4 ], [ 4, 5 ], [ 4, 53 ], ...]}
55
+ ```
56
+
57
+ The average for the `seq` and the `label` are provided below:
58
+
59
+ | Feature | Mean Count |
60
+ | ---------- | ---------------- |
61
+ | seq | 249 |
62
+ | label | 1,500 |
63
+
64
+ ### Data Fields
65
+
66
+ - `seq`: a string containing the protein seqeunce
67
+ - `label`: a string containing the contact label of each residue pair.
68
+
69
+ ### Data Splits
70
+
71
+ The CNN/DailyMail dataset has 3 splits: _train_, _validation_, and _test_. Below are the statistics for Version 3.0.0 of the dataset.
72
+
73
+ | Dataset Split | Number of Instances in Split |
74
+ | ------------- | ------------------------------------------- |
75
+ | Train | 12,041 |
76
+ | Validation | 1,505 |
77
+ | Test | 1,505 |
78
+
79
+ ### Source Data
80
+
81
+ #### Initial Data Collection and Normalization
82
+
83
+ The [trRosetta dataset]() is employed and the same split.
84
+
85
+
86
+ ### Licensing Information
87
+
88
+ The dataset is released under the [Apache-2.0 License](http://www.apache.org/licenses/LICENSE-2.0).
89
+
90
+ ### Citation
91
+ If you find our work useful, please consider citing the following paper:
92
+
93
+ ```
94
+ @misc{chen2024xtrimopglm,
95
+ title={xTrimoPGLM: unified 100B-scale pre-trained transformer for deciphering the language of protein},
96
+ 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},
97
+ year={2024},
98
+ eprint={2401.06199},
99
+ archivePrefix={arXiv},
100
+ primaryClass={cs.CL},
101
+ note={arXiv preprint arXiv:2401.06199}
102
+ }
103
+ ```
104
+
105