Add dataset card

README.md

@@ -1,37 +1,87 @@
 ---
-dataset_info:
-  features:
-  - name: AID
-    dtype: int64
-  - name: CID
-    dtype: int64
-  - name: SMILES
-    dtype: string
-  - name: InChI
-    dtype: string
-  - name: InChIKey
-    dtype: string
-  - name: SMILES_sanitized
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 1784959
-    num_examples: 6398
-  - name: test
-    num_bytes: 198359
-    num_examples: 711
-  - name: full
-    num_bytes: 1983319
-    num_examples: 7109
-  download_size: 2007452
-  dataset_size: 3966637
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
-  - split: test
-    path: data/test-*
-  - split: full
-    path: data/full-*
+license: cc-by-4.0
+task_categories:
+  - tabular-classification
+tags:
+  - chemistry
+  - drug-discovery
+  - pubchem
+  - bioassay
+  - smiles
+  - rdkit
 ---
+
+# Simeonov2008: PubChem Bioassay Active Compounds
+
+## Overview
+
+Active compounds from PubChem qHTS bioassay screenings (AIDs: 587, 588, 590, 591, 592, 593, 594) based on Simeonov et al. (2008). SMILES were sanitized using RDKit MolStandardize.
+
+| Metric | Value |
+|--------|-------|
+| Total rows | 7109 |
+| Unique compounds | 6743 |
+| Assays | 6 |
+
+## Quick Start
+
+```python
+from datasets import load_dataset
+
+# Load dataset
+dataset = load_dataset("jpraise/Simeonov2008")
+
+# Get full data as DataFrame
+df = dataset['full'].to_pandas()
+print(df.head())
+```
+
+## Data Fields
+
+| Column | Description |
+|--------|-------------|
+| AID | PubChem Assay ID |
+| CID | PubChem Compound ID |
+| SMILES | Original SMILES from PubChem |
+| InChI | IUPAC InChI identifier |
+| InChIKey | Hashed InChI key |
+| SMILES_sanitized | Standardized SMILES via RDKit |
+
+## Assays
+
+| AID | Link |
+|-----|------|
+| 587 | [PubChem](https://pubchem.ncbi.nlm.nih.gov/bioassay/587) |
+| 588 | [PubChem](https://pubchem.ncbi.nlm.nih.gov/bioassay/588) |
+| 590 | [PubChem](https://pubchem.ncbi.nlm.nih.gov/bioassay/590) |
+| 591 | [PubChem](https://pubchem.ncbi.nlm.nih.gov/bioassay/591) |
+| 592 | [PubChem](https://pubchem.ncbi.nlm.nih.gov/bioassay/592) |
+| 593 | [PubChem](https://pubchem.ncbi.nlm.nih.gov/bioassay/593) |
+| 594 | [PubChem](https://pubchem.ncbi.nlm.nih.gov/bioassay/594) |
+
+## SMILES Sanitization
+
+RDKit MolStandardize pipeline:
+1. LargestFragmentChooser - Remove salts/counterions
+2. Normalizer - Standardize functional groups
+3. Uncharger - Neutralize charges
+4. TautomerEnumerator - Canonicalize tautomers
+
+## Citation
+
+```bibtex
+@article{simeonov2008,
+  title={Fluorescence spectroscopic profiling of compound libraries},
+  author={Simeonov, Anton and Jadhav, Ajit and Thomas, Craig J and others},
+  journal={Journal of Medicinal Chemistry},
+  volume={51},
+  number={8},
+  pages={2363--2371},
+  year={2008},
+  doi={10.1021/jm701301m}
+}
+```
+
+## License
+
+CC-BY-4.0. Source data from PubChem (public domain).