metadata
pretty_name: Simeonov2008 Bioactive Compounds
tags:
- bioassay
- chemical
- small-molecule
- SMILES
- PubChem
language: en
license: unknown
dataset_info:
features:
- name: AID
dtype: int64
- name: CID
dtype: int64
- name: SMILES
dtype: large_string
- name: InChI
dtype: large_string
- name: InChIKey
dtype: large_string
- name: SMILES_sanitized
dtype: large_string
splits:
- name: train
num_bytes: 2108290
num_examples: 7109
download_size: 977966
dataset_size: 2108290
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
Simeonov2008 Bioactive Compounds
Dataset Description
This dataset contains small-molecule compounds tested in multiple PubChem bioassays (AID 587, 588, 590, 591, 593, 594). It includes canonical SMILES, sanitized SMILES, InChI, and InChIKey identifiers for 7,109 unique compounds. Sanitization was performed to standardize chemical structures, remove counterions, and normalize functional groups.
Citation:
Simeonov, A., et al. (2008). High-throughput screening for bioactive compounds. PubChem BioAssay Database. https://pubchem.ncbi.nlm.nih.gov/
Quick-Start Guide
from datasets import load_dataset
# Load the dataset from Hugging Face
dataset = load_dataset("adsfibonacci/Simeonov2008")
df = dataset['train'].to_pandas()
# Preview
print(df.head())