Stage: Lead ID / ADMET
24 benchmarks, ranked by composite score. ← All stages
| # | Benchmark | Description | Modalities | Score | Flags |
|---|---|---|---|---|---|
| 1 | TDC ADMET Group | 22-task ADMET benchmark suite with scaffold splits — core TDC leaderboard. | small-molecule | 100.0 | |
| 2 | SAbDab | Structural antibody database — curated PDB antibody structures with annotation. | biologic-mab | 100.0 | |
| 3 | ChEMBL | Manually curated bioactive molecule DB; backbone for most ML chemistry benchmarks. | small-moleculeprotein-general | 97.5 | |
| 4 | Observed Antibody Space (OAS) | Repository of >1B antibody BCR sequences from public repertoires — core for antibody LM pretraining. | biologic-mab | 97.5 | |
| 5 | ProteinGym | 217 DMS substitution assays + indel + clinical variants — de facto standard for VEPs. | protein-general | 97.5 | |
| 6 | Practical Molecular Optimization (PMO) | Budget-aware benchmark for goal-directed molecule optimization with 23 oracle functions. | small-molecule | 88.9 | |
| 7 | Polaris ADMET | Industry-contributed ADMET benchmarks on Polaris Hub (Novartis/AstraZeneca/Polaris SC curated endpoints). | small-molecule | 88.4 | |
| 8 | ISM Benchmarks: GPCRs (Insilico) | 87-benchmark GPCR affinity suite on ScienceAIBench / InsilicoBench / DDB — kinome-scale coverage of class A/B/C GPCRs. | small-moleculeprotein-general | 87.6 | |
| 9 | ToxCast | EPA's in vitro toxicity screening dataset — ~700 assays × ~9k chemicals. | small-molecule | 85.6 | |
| 10 | ISM Benchmarks: ADMET (Insilico) | 28-endpoint ADMET benchmark suite on ScienceAIBench/InsilicoBench/DDB. | small-molecule | 84.6 | |
| 11 | MoleculeACE | Benchmark testing model robustness on activity cliffs across 30 ChEMBL targets. | small-molecule | 83.3 | |
| 12 | GuacaMol | Goal-directed + distribution-learning benchmarks for molecular generative models. | small-molecule | 80.5 | |
| 13 | ADMET-AI | Graph-based ADMET prediction + benchmark on 41 endpoints leveraging ChEMBL + TDC. | small-molecule | 79.5 | |
| 14 | AMES (mutagenicity) | AMES bacterial mutagenicity benchmark — standard gentox endpoint. | small-molecule | 79.5 | |
| 15 | MoleculeNet | Multi-task molecular ML benchmark covering quantum, physical, biophysical, physiological properties. | small-molecule | 78.0 | data-leakage-known |
| 16 | USPTO-50K / USPTO-MIT (Retrosynthesis) | Reactions extracted from USPTO patents; standard retrosynthesis/forward-reaction benchmark. | small-molecule | 78.0 | data-leakage-known |
| 17 | Tox21 | US Tox21 program HTS data on 12 nuclear receptor / stress response assays. | small-molecule | 77.5 | |
| 18 | TDC DrugSyn (OncoPolyPharm + DrugComb_NCI60) | Drug combination synergy benchmark (NCI ALMANAC + OncoPolyPharmacology). | small-molecule | 77.0 | |
| 19 | Obach PK Dataset | Obach human PK dataset (t1/2, VDss, CL) — standard human-PK ML benchmark. | small-molecule | 77.0 | |
| 20 | PDBbind | Curated PDB complexes with experimental binding affinities — de facto standard for ML affinity prediction. | small-moleculeprotein-general | 75.9 | data-leakage-known |
| 21 | hERG (cardio-tox) TDC | Cardiac tox benchmark (hERG inhibition) — standardized from Wang et al. | small-molecule | 73.9 | |
| 22 | DILI / LD50 Zhu | Drug-induced liver injury + rat LD50 (Zhu) — standard acute tox benchmarks. | small-molecule | 73.9 | |
| 23 | MOSES | Molecular Sets — distribution learning benchmark with 8 reference metrics on ZINC subset. | small-molecule | 72.4 | |
| 24 | ClinTox | Binary classification of FDA-approved vs. trial-failed-for-toxicity compounds. | small-molecule | 65.6 | data-leakage-known |