Benchmark Datasets
MoleculeNet ADMET, toxicity and solubility tasks, a large virtual-screening set, and medical QA for the biomedical models.
Viewer • Updated • 2.04k • 47Note Blood-brain barrier penetration, 2,039 molecules - binary classification
Aurigene-AI/MoleculeNet_BACE
Viewer • Updated • 1.51k • 51Note BACE-1 inhibition (Alzheimer's target), 1,513 compounds
Aurigene-AI/MoleculeNet_ClinTox
Viewer • Updated • 1.48k • 51Note Clinical trial toxicity and FDA approval status, 1,477 drugs
Aurigene-AI/MoleculeNet_Tox21
Viewer • Updated • 7.83k • 65Note 12 nuclear-receptor and stress-response assays, 7,831 compounds
Aurigene-AI/MoleculeNet_SIDER
Viewer • Updated • 1.43k • 46Note Marketed-drug side effects by MedDRA organ class, 1,427 drugs
Aurigene-AI/MoleculeNet_ESOL
Viewer • Updated • 1.13k • 44Note Aqueous solubility regression, 1,128 compounds
Aurigene-AI/MoleculeNet_FreeSolv
Viewer • Updated • 642 • 46Note Hydration free energy regression, 642 molecules
Aurigene-AI/MoleculeNet_Lipophilicity
Viewer • Updated • 4.2k • 53Note logD at pH 7.4, 4,200 compounds - core ADMET endpoint
Aurigene-AI/MoleculeNet_HIV
Viewer • Updated • 41.1k • 57Note 41,127 compounds screened for HIV replication inhibition
Aurigene-AI/MedQA-USMLE-4-options
Viewer • Updated • 11.5k • 61Note US medical licensing exam questions, 11,451 items
Aurigene-AI/MedMCQA
Viewer • Updated • 193k • 52Note Medical entrance exam questions, 193,155 items