eacortes commited on
Commit
5728221
·
1 Parent(s): 551b635

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +74 -17
README.md CHANGED
@@ -1,19 +1,76 @@
1
  ---
2
- dataset_info:
3
- features:
4
- - name: smiles
5
- dtype: string
6
- - name: source_id
7
- dtype: int64
8
- splits:
9
- - name: train
10
- num_bytes: 952803468
11
- num_examples: 3474464
12
- download_size: 283961210
13
- dataset_size: 952803468
14
- configs:
15
- - config_name: default
16
- data_files:
17
- - split: train
18
- path: data/train-*
19
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ license: cc-by-4.0
3
+ language:
4
+ - en
5
+ tags:
6
+ - chemistry
7
+ - molecules
8
+ - reactions
9
+ - smiles
10
+ - cheminformatics
11
+ size_categories:
12
+ - 1M<n<10M
13
+ task_categories:
14
+ - text-generation
15
+ - fill-mask
16
+ pretty_name: Reaction SMILES Dataset
 
 
17
  ---
18
+
19
+ # Reaction SMILES Dataset
20
+
21
+ A curated dataset of **3,474,464 unique chemical reaction SMILES strings** compiled from multiple public sources for use in pre-training and fine-tuning molecular language models.
22
+
23
+ ## Dataset Description
24
+
25
+ ### What are Reaction SMILES?
26
+
27
+ Reaction SMILES (Simplified Molecular Input Line Entry System) extend the standard SMILES notation to represent complete chemical reactions. They encode reactants, reagents/catalysts, and products in a single text string using the `>` delimiter:
28
+
29
+ ```
30
+ reactants>reagents>products
31
+ ```
32
+
33
+ For example: `CC(=O)O.CCO>>[H+]>CC(=O)OCC.O` represents the esterification of acetic acid with ethanol to form ethyl acetate.
34
+
35
+ ### Dataset Features
36
+
37
+ - **smiles**: The reaction SMILES string
38
+ - **source_id**: Integer identifier linking to the original data source
39
+
40
+ ## Data Sources
41
+
42
+ | Source ID | Source URL |
43
+ | --------- | -------------------------------------------------------------------------------------------------------------------------------------- |
44
+ | 1 | [US Patents 1976-Sep2016 Grants](https://figshare.com/articles/dataset/Chemical_reactions_from_US_patents_1976-Sep2016_/5104873) |
45
+ | 2 | [US Patents 2001-Sep2016 Applications](https://figshare.com/articles/dataset/Chemical_reactions_from_US_patents_1976-Sep2016_/5104873) |
46
+ | 3 | [CRD 1.37M Dataset (2024)](https://figshare.com/articles/dataset/Reaction_SMILES_b_CRD_b_1_37M_dataset/28230053/1) |
47
+ | 4 | [USPTO Year 2023](https://figshare.com/articles/dataset/Reaction_SMILES_USPTO_year_2023/24921555) |
48
+ | 5 | [Reaction SMILES Dataset (2023)](https://figshare.com/articles/dataset/Reaction_SMILES_dataset/22491730) |
49
+
50
+ ## Dataset Statistics
51
+
52
+ | Metric | Value |
53
+ | --------------------------------------- | --------- |
54
+ | Total unique reactions | 3,474,464 |
55
+ | Original entries (before deduplication) | 5,992,346 |
56
+ | Duplicates removed | 2,517,882 |
57
+
58
+ ## Processing
59
+
60
+ The dataset was created by:
61
+ 1. Loading all source files
62
+ 2. Extracting reaction SMILES strings
63
+ 3. Stripping whitespace
64
+ 4. Removing duplicates (keeping first occurrence to preserve source attribution)
65
+
66
+ ## Usage
67
+
68
+ ```python
69
+ from datasets import load_dataset
70
+
71
+ dataset = load_dataset("Derify/ReactionSmiles")
72
+ ```
73
+
74
+ ## License
75
+
76
+ This dataset aggregates publicly available reaction data. Please refer to the individual source links for specific licensing terms.