Datasets:
Commit
·
8d2584e
1
Parent(s):
7a00104
basic details
Browse filesadded reference from the December 1.0 paper along with the molecule selection filter
README.md
CHANGED
@@ -1,3 +1,26 @@
|
|
1 |
---
|
2 |
license: openrail
|
|
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: openrail
|
3 |
+
pretty_name: PubChem 68K
|
4 |
+
size_categories:
|
5 |
+
- 10K<n<100K
|
6 |
+
task_categories:
|
7 |
+
- image-to-text
|
8 |
---
|
9 |
+
|
10 |
+
Molecules in this set
|
11 |
+
|
12 |
+
* have a molecular weight of fewer than 1500 Daltons,
|
13 |
+
* not possess counter ions,
|
14 |
+
* only contain the elements C, H, O, N, P, S, F, Cl, Br, I, Se and B,
|
15 |
+
* not contain isotopes of Hydrogens (D, T),
|
16 |
+
* have 3–40 bonds,
|
17 |
+
* not contain any charged groups including zwitterionic forms,
|
18 |
+
* only contain implicit hydrogens, except in functional groups,
|
19 |
+
* have less than 40 SMILES characters,
|
20 |
+
* no stereochemistry is allowed.
|
21 |
+
|
22 |
+
The original dataset from Decimer was imported and randomly sampled. 516x516 sized images were generated using RDKit.
|
23 |
+
|
24 |
+
## Reference
|
25 |
+
|
26 |
+
> Rajan, Kohulan; Zielesny, Achim; Steinbeck, Christoph (2021): DECIMER 1.0: Deep Learning for Chemical Image Recognition using Transformers. ChemRxiv. Preprint. https://doi.org/10.26434/chemrxiv.14479287.v1
|