Drug Discovery Benchmarks (DDB)
Insilico's drug-discovery-specific benchmark portal: TargetBench, Longevity Benchmark, GPCR affinity, PDBbind-style tasks, ISM ADMET, TDC ADMET mirror, ClinBench, biologics.
Kind
meta-platform
Composite
87.6
Benchmarks tracked
206
Direct-linked
206
As of
2026-05-12
Host organisation
Insilico Medicine
Primary contacts
Alex Zhavoronkov, Alex Aliper
Founded
2025
License model
CC-BY (per portal)
Official site
GitHub
→ GitHub
Count methodology
Fetched https://ddb.insilico.com/api/benchmarks on 2026-05-12; meta.totalBenchmarks=206 across 6 categories × 15 suites.
Rubric
rigor
4
coverage
5
maintenance
5
adoption
3
quality
4
accessibility
5
industry_relevance
5
Breakdown
Biology (TargetBench + Longevity)
29
Affinity and Binding
94
Chemical Synthesis
2
ADMET, PK & Safety
50
Clinical Trials
25
Biologics
6
Notes
Drug-discovery focused cut. Includes a mirror of TDC ADMET for cross-platform comparability.
Hosted benchmarks (206)
Direct-linked — each links to the benchmark’s leaderboard / detail page on the host portal.
ADMET, PK & Safety (50)
BCRP InhibitionISM Benchmarks: ADMET · Breast Cancer Resistance Protein (BCRP/ABCG2) inhibition prediction. BCRP is an efflux transporter affecting oral absorpCaco-2 Efflux RatioISM Benchmarks: ADMET · Caco-2 efflux ratio measures the ratio of basolateral-to-apical vs apical-to-basolateral transport, indicating active efCaco-2 PermeabilityISM Benchmarks: ADMET · Caco-2 cell effective permeability (Papp) using human colon carcinoma cells to model passive diffusion and active transpClearance Human HepatocytesISM Benchmarks: ADMET · Intrinsic clearance in human hepatocytes captures both Phase I (CYP-mediated) and Phase II (conjugation) metabolism. ProClearance Human Liver MicrosomesISM Benchmarks: ADMET · Intrinsic clearance in human liver microsomes (HLM) evaluates CYP-mediated oxidative metabolism. A standard early-stage Clearance Mouse Liver MicrosomesISM Benchmarks: ADMET · Intrinsic clearance in mouse liver microsomes (MLM). Mouse is a common preclinical species; predicting species-specific Clearance Rat Liver MicrosomesISM Benchmarks: ADMET · Intrinsic clearance in rat liver microsomes (RLM). Rat is the most widely used preclinical species for PK studies. ProprCYP1A2 Inhibition IC50ISM Benchmarks: ADMET · CYP1A2 inhibition potency (IC50) prediction. CYP1A2 metabolizes several important drugs including caffeine and theophyllCYP2C19 Inhibition IC50ISM Benchmarks: ADMET · CYP2C19 inhibition potency (IC50) prediction. CYP2C19 is a highly polymorphic enzyme metabolizing proton pump inhibitorsCYP2C9 Inhibition IC50ISM Benchmarks: ADMET · CYP2C9 inhibition potency (IC50) prediction. CYP2C9 metabolizes ~15% of clinical drugs including warfarin and NSAIDs; inCYP2D6 Inhibition IC50ISM Benchmarks: ADMET · CYP2D6 inhibition potency (IC50) prediction. CYP2D6 metabolizes ~25% of marketed drugs and is highly polymorphic in the CYP3A4 Inhibition IC50ISM Benchmarks: ADMET · CYP3A4 inhibition potency (IC50) prediction. CYP3A4 is the most abundant hepatic CYP, metabolizing ~50% of marketed drugHalf Life HumanISM Benchmarks: ADMET · Human plasma half-life (T1/2) in hours. Determines how long a drug remains active in the body and directly impacts dosinHEK293 CC50ISM Benchmarks: ADMET · CC50 (cytotoxic concentration 50%) in HEK293 cells measures the concentration required to reduce cell viability by 50%. HepG2 CC50ISM Benchmarks: ADMET · CC50 (cytotoxic concentration 50%) in HepG2 hepatocellular carcinoma cells. A key indicator of hepatotoxicity risk and ghERG IC50ISM Benchmarks: ADMET · hERG potassium channel inhibition potency (IC50) prediction. hERG inhibition causes QT prolongation and potentially fataKinetic SolubilityISM Benchmarks: ADMET · Kinetic solubility measures the concentration at which a compound precipitates from a DMSO stock solution in aqueous bufLogDISM Benchmarks: ADMET · LogD (distribution coefficient at pH 7.4) measures the ratio of a compound's concentration in octanol vs aqueous phase aLogPISM Benchmarks: ADMET · LogP (partition coefficient) measures the lipophilicity of a neutral compound between octanol and water. A fundamental pMDCK-MDR1 Efflux RatioISM Benchmarks: ADMET · MDCK-MDR1 efflux ratio measures P-gp-mediated efflux using MDCK cells overexpressing MDR1. High efflux ratios indicate PMDCK-MDR1 PermeabilityISM Benchmarks: ADMET · MDCK-MDR1 apparent permeability (Papp) measures drug transport across MDCK cells expressing the MDR1 P-glycoprotein tranP-gp InhibitionISM Benchmarks: ADMET · P-glycoprotein (P-gp/MDR1) inhibition prediction. P-gp is a major efflux transporter; its inhibition can alter pharmacokPAMPA PermeabilityISM Benchmarks: ADMET · Parallel Artificial Membrane Permeability Assay (PAMPA) provides a cell-free, high-throughput measure of passive transcePPB HumanISM Benchmarks: ADMET · Plasma protein binding (PPB) in human plasma. Only the unbound fraction of a drug can cross membranes and engage its tarPPB MouseISM Benchmarks: ADMET · Plasma protein binding (PPB) in mouse plasma. Species-specific binding differences are essential for translating preclinPPB RatISM Benchmarks: ADMET · Plasma protein binding (PPB) in rat plasma. Species-specific binding is critical for PK translation from the most commonThermodynamic SolubilityISM Benchmarks: ADMET · Thermodynamic solubility is the equilibrium solubility of the most stable crystalline form in aqueous media. A definitivVDssISM Benchmarks: ADMET · Volume of distribution at steady state (VDss) measures how extensively a drug distributes from plasma into body tissues.AMES MutagenicityTDC ADMET · Ames mutagenicity test classification predicting whether a compound is mutagenic. The Ames test is a regulatory requiremBBB PenetrationTDC ADMET · Blood-brain barrier penetration classification predicting whether a compound can cross the BBB. Essential for CNS drug dBioavailabilityTDC ADMET · Oral bioavailability classification predicting whether a drug achieves sufficient systemic exposure after oral administrCaco-2 PermeabilityTDC ADMET · Caco-2 cell effective permeability measures a compound's ability to cross the intestinal epithelial barrier via passive Clearance HepatocyteTDC ADMET · Intrinsic clearance measured in hepatocytes (µL/min/10^6 cells). Provides a more physiologically complete assessment of Clearance MicrosomeTDC ADMET · Intrinsic clearance measured in liver microsomes (mL/min/g). Reflects the rate of drug metabolism by microsomal enzymes,CYP2C9 InhibitionTDC ADMET · CYP2C9 inhibition classification predicting whether a compound inhibits the CYP2C9 enzyme. CYP2C9 metabolizes ~15% of clCYP2C9 SubstrateTDC ADMET · CYP2C9 substrate classification predicting whether a compound is metabolized by the CYP2C9 enzyme. Important for predictCYP2D6 InhibitionTDC ADMET · CYP2D6 inhibition classification predicting whether a compound inhibits the CYP2D6 enzyme. CYP2D6 metabolizes ~25% of maCYP2D6 SubstrateTDC ADMET · CYP2D6 substrate classification predicting whether a compound is metabolized by the CYP2D6 enzyme. CYP2D6 is highly polyCYP3A4 InhibitionTDC ADMET · CYP3A4 inhibition classification predicting whether a compound inhibits the CYP3A4 enzyme. Given CYP3A4's dominant role CYP3A4 SubstrateTDC ADMET · CYP3A4 substrate classification predicting whether a compound is metabolized by the CYP3A4 enzyme. CYP3A4 is the most abDILITDC ADMET · Drug-Induced Liver Injury (DILI) classification predicting whether a drug causes liver damage. DILI is a leading cause oHalf LifeTDC ADMET · Drug half-life (hours) measures the time required for the plasma concentration to decrease by 50%. Critical for determinhERG BlockersTDC ADMET · hERG channel blocker classification predicting whether a compound inhibits the human ether-à-go-go-related gene potassiuHIATDC ADMET · Human intestinal absorption (HIA) classification predicting whether a drug is absorbed through the human intestine (>80%LD50 Acute ToxicityTDC ADMET · Acute toxicity prediction as LD50 in log(1/(mol/kg)). LD50 is the dose required to kill 50% of test animals and is a funLipophilicityTDC ADMET · Experimental lipophilicity measured as octanol/water distribution coefficient (logD at pH 7.4). A key physicochemical prP-gp InhibitionTDC ADMET · P-glycoprotein (P-gp) inhibition classification. P-gp is an efflux transporter that limits oral drug absorption and braiPPBRTDC ADMET · Plasma protein binding rate (%) from AstraZeneca. Only the unbound fraction of a drug is pharmacologically active; high SolubilityTDC ADMET · Aqueous solubility prediction in log mol/L. Solubility directly impacts oral bioavailability and is a critical parameterVDssTDC ADMET · Volume of distribution at steady state (L/kg) measures how extensively a drug distributes from plasma into body tissues.
Affinity and Binding (94)
5-HT1A IC50GPCR Affinity Suite · GPCR 5-HT1A IC50 affinity prediction (IC50 or Ki).5-HT1A KIGPCR Affinity Suite · GPCR 5-HT1A KI affinity prediction (IC50 or Ki).5-HT1C KIGPCR Affinity Suite · GPCR 5-HT1C KI affinity prediction (IC50 or Ki).5-HT2A IC50GPCR Affinity Suite · GPCR 5-HT2A IC50 affinity prediction (IC50 or Ki).5-HT2A KIGPCR Affinity Suite · GPCR 5-HT2A KI affinity prediction (IC50 or Ki).5-HT2B KIGPCR Affinity Suite · GPCR 5-HT2B KI affinity prediction (IC50 or Ki).5-HT5A KIGPCR Affinity Suite · GPCR 5-HT5A KI affinity prediction (IC50 or Ki).5-HT6 IC50GPCR Affinity Suite · GPCR 5-HT6 IC50 affinity prediction (IC50 or Ki).5-HT6 KIGPCR Affinity Suite · GPCR 5-HT6 KI affinity prediction (IC50 or Ki).5-HT7 KIGPCR Affinity Suite · GPCR 5-HT7 KI affinity prediction (IC50 or Ki).A1 IC50GPCR Affinity Suite · GPCR A1 IC50 affinity prediction (IC50 or Ki).A1 KIGPCR Affinity Suite · GPCR A1 KI affinity prediction (IC50 or Ki).A2A IC50GPCR Affinity Suite · GPCR A2A IC50 affinity prediction (IC50 or Ki).A2A KIGPCR Affinity Suite · GPCR A2A KI affinity prediction (IC50 or Ki).A2B IC50GPCR Affinity Suite · GPCR A2B IC50 affinity prediction (IC50 or Ki).A2B KIGPCR Affinity Suite · GPCR A2B KI affinity prediction (IC50 or Ki).A3 KIGPCR Affinity Suite · GPCR A3 KI affinity prediction (IC50 or Ki).ALPHA1A KIGPCR Affinity Suite · GPCR ALPHA1A KI affinity prediction (IC50 or Ki).B1 IC50GPCR Affinity Suite · GPCR B1 IC50 affinity prediction (IC50 or Ki).B1 KIGPCR Affinity Suite · GPCR B1 KI affinity prediction (IC50 or Ki).CASR IC50GPCR Affinity Suite · GPCR CASR IC50 affinity prediction (IC50 or Ki).CB1 IC50GPCR Affinity Suite · GPCR CB1 IC50 affinity prediction (IC50 or Ki).CB1 KIGPCR Affinity Suite · GPCR CB1 KI affinity prediction (IC50 or Ki).CB2 IC50GPCR Affinity Suite · GPCR CB2 IC50 affinity prediction (IC50 or Ki).CB2 KIGPCR Affinity Suite · GPCR CB2 KI affinity prediction (IC50 or Ki).CCR1 IC50GPCR Affinity Suite · GPCR CCR1 IC50 affinity prediction (IC50 or Ki).CCR2 IC50GPCR Affinity Suite · GPCR CCR2 IC50 affinity prediction (IC50 or Ki).CCR3 IC50GPCR Affinity Suite · GPCR CCR3 IC50 affinity prediction (IC50 or Ki).CCR5 IC50GPCR Affinity Suite · GPCR CCR5 IC50 affinity prediction (IC50 or Ki).CRHR1 IC50GPCR Affinity Suite · GPCR CRHR1 IC50 affinity prediction (IC50 or Ki).CRHR1 KIGPCR Affinity Suite · GPCR CRHR1 KI affinity prediction (IC50 or Ki).CXCR3 IC50GPCR Affinity Suite · GPCR CXCR3 IC50 affinity prediction (IC50 or Ki).CXCR4 IC50GPCR Affinity Suite · GPCR CXCR4 IC50 affinity prediction (IC50 or Ki).D1 KIGPCR Affinity Suite · GPCR D1 KI affinity prediction (IC50 or Ki).D2 IC50GPCR Affinity Suite · GPCR D2 IC50 affinity prediction (IC50 or Ki).D2 KIGPCR Affinity Suite · GPCR D2 KI affinity prediction (IC50 or Ki).D3 KIGPCR Affinity Suite · GPCR D3 KI affinity prediction (IC50 or Ki).D4 KIGPCR Affinity Suite · GPCR D4 KI affinity prediction (IC50 or Ki).EP2 IC50GPCR Affinity Suite · GPCR EP2 IC50 affinity prediction (IC50 or Ki).EP4 IC50GPCR Affinity Suite · GPCR EP4 IC50 affinity prediction (IC50 or Ki).ETA IC50GPCR Affinity Suite · GPCR ETA IC50 affinity prediction (IC50 or Ki).GHS-R IC50GPCR Affinity Suite · GPCR GHS-R IC50 affinity prediction (IC50 or Ki).GLUCAGON IC50GPCR Affinity Suite · GPCR GLUCAGON IC50 affinity prediction (IC50 or Ki).GLUCAGON KIGPCR Affinity Suite · GPCR GLUCAGON KI affinity prediction (IC50 or Ki).GNRHR IC50GPCR Affinity Suite · GPCR GNRHR IC50 affinity prediction (IC50 or Ki).GNRHR KIGPCR Affinity Suite · GPCR GNRHR KI affinity prediction (IC50 or Ki).GPR44 IC50GPCR Affinity Suite · GPCR GPR44 IC50 affinity prediction (IC50 or Ki).GPR44 KIGPCR Affinity Suite · GPCR GPR44 KI affinity prediction (IC50 or Ki).IL8RB IC50GPCR Affinity Suite · GPCR IL8RB IC50 affinity prediction (IC50 or Ki).MACHRM1 IC50GPCR Affinity Suite · GPCR MACHRM1 IC50 affinity prediction (IC50 or Ki).MACHRM1 KIGPCR Affinity Suite · GPCR MACHRM1 KI affinity prediction (IC50 or Ki).MACHRM2 KIGPCR Affinity Suite · GPCR MACHRM2 KI affinity prediction (IC50 or Ki).MACHRM3 KIGPCR Affinity Suite · GPCR MACHRM3 KI affinity prediction (IC50 or Ki).MACHRM4 IC50GPCR Affinity Suite · GPCR MACHRM4 IC50 affinity prediction (IC50 or Ki).MACHRM4 KIGPCR Affinity Suite · GPCR MACHRM4 KI affinity prediction (IC50 or Ki).MC3R KIGPCR Affinity Suite · GPCR MC3R KI affinity prediction (IC50 or Ki).MC4R IC50GPCR Affinity Suite · GPCR MC4R IC50 affinity prediction (IC50 or Ki).MC4R KIGPCR Affinity Suite · GPCR MC4R KI affinity prediction (IC50 or Ki).MC5R KIGPCR Affinity Suite · GPCR MC5R KI affinity prediction (IC50 or Ki).MGLUR1 IC50GPCR Affinity Suite · GPCR MGLUR1 IC50 affinity prediction (IC50 or Ki).MGLUR2 IC50GPCR Affinity Suite · GPCR MGLUR2 IC50 affinity prediction (IC50 or Ki).MGLUR5 IC50GPCR Affinity Suite · GPCR MGLUR5 IC50 affinity prediction (IC50 or Ki).MGLUR5 KIGPCR Affinity Suite · GPCR MGLUR5 KI affinity prediction (IC50 or Ki).NK1 KIGPCR Affinity Suite · GPCR NK1 KI affinity prediction (IC50 or Ki).NK2 IC50GPCR Affinity Suite · GPCR NK2 IC50 affinity prediction (IC50 or Ki).NK2 KIGPCR Affinity Suite · GPCR NK2 KI affinity prediction (IC50 or Ki).NK3 IC50GPCR Affinity Suite · GPCR NK3 IC50 affinity prediction (IC50 or Ki).NK3 KIGPCR Affinity Suite · GPCR NK3 KI affinity prediction (IC50 or Ki).NPY1R KIGPCR Affinity Suite · GPCR NPY1R KI affinity prediction (IC50 or Ki).NPY2R IC50GPCR Affinity Suite · GPCR NPY2R IC50 affinity prediction (IC50 or Ki).NPY5R IC50GPCR Affinity Suite · GPCR NPY5R IC50 affinity prediction (IC50 or Ki).NPY5R KIGPCR Affinity Suite · GPCR NPY5R KI affinity prediction (IC50 or Ki).OPRD1 IC50GPCR Affinity Suite · GPCR OPRD1 IC50 affinity prediction (IC50 or Ki).OPRD1 KIGPCR Affinity Suite · GPCR OPRD1 KI affinity prediction (IC50 or Ki).OPRK1 IC50GPCR Affinity Suite · GPCR OPRK1 IC50 affinity prediction (IC50 or Ki).OPRK1 KIGPCR Affinity Suite · GPCR OPRK1 KI affinity prediction (IC50 or Ki).OPRM1 IC50GPCR Affinity Suite · GPCR OPRM1 IC50 affinity prediction (IC50 or Ki).OPRM1 KIGPCR Affinity Suite · GPCR OPRM1 KI affinity prediction (IC50 or Ki).OX1 IC50GPCR Affinity Suite · GPCR OX1 IC50 affinity prediction (IC50 or Ki).OX1 KIGPCR Affinity Suite · GPCR OX1 KI affinity prediction (IC50 or Ki).OX2 IC50GPCR Affinity Suite · GPCR OX2 IC50 affinity prediction (IC50 or Ki).OX2 KIGPCR Affinity Suite · GPCR OX2 KI affinity prediction (IC50 or Ki).THROMBIN IC50GPCR Affinity Suite · GPCR THROMBIN IC50 affinity prediction (IC50 or Ki).V1A IC50GPCR Affinity Suite · GPCR V1A IC50 affinity prediction (IC50 or Ki).V1A KIGPCR Affinity Suite · GPCR V1A KI affinity prediction (IC50 or Ki).V2 IC50GPCR Affinity Suite · GPCR V2 IC50 affinity prediction (IC50 or Ki).V2 KIGPCR Affinity Suite · GPCR V2 KI affinity prediction (IC50 or Ki).IC50 Prediction (Cold Drug)IC50 prediction · Evaluates IC50 (half-maximal inhibitory concentration) prediction using BindingDB dataset with Cold Drug split. IC50 is MMFF Energy PredictionMolecular Mechanics Energy · Predicting Molecular Mechanics (MM) energy for diverse 3D conformations. A computationally designed drug is useless if iNumber of Non-Covalent InteractionsPDBBind Number of Interactions · Evaluates prediction of the number of non-covalent interactions in protein-ligand complexes using LP-PDBBind dataset. Gr3D Interactions RestorationPDBBind Restore Interactions · Evaluates the ability to restore 3D protein-ligand interactions from LP-PDBBind dataset. Measures the ratio of restored PKIS2 EGFRPKIS2 Kinase Inhibition Suite · Kinase inhibition prediction for EGFR (Published Kinase Inhibitor Set 2). Essential for designing selective kinase inhibPKIS2 KITPKIS2 Kinase Inhibition Suite · Kinase inhibition prediction for KIT (Published Kinase Inhibitor Set 2). Essential for designing selective kinase inhibiPKIS2 RETPKIS2 Kinase Inhibition Suite · Kinase inhibition prediction for RET (Published Kinase Inhibitor Set 2). Predicting how small molecules interact with th
Biologics (6)
CHO PolyreactivityPolyreactivity · Evaluates antibody binding to CHO-cell membrane proteins (CHO) as a measure of polyreactivity. Polyreactivity is the proOvalbumin PolyreactivityPolyreactivity · Evaluates antibody binding to Ovalbumin as a measure of polyreactivity. Polyreactivity is the propensity of an antibody SEC %MonomerSEC SMAC · SEC %Monomer, assessed by size-exclusion chromatography, reflects the proportion of monomeric antibody and serves as an SMACSEC SMAC · SMAC measures antibody retention time using standup monolayer affinity chromatography, where longer retention reflects sTiterTm and Titer · Evaluates titer using Spearman correlation.TmTm and Titer · Evaluates Tm using Spearman correlation.
Biology (29)
Target Identification - CancerTargetBench · Evaluation of AI models on target identification for Cancer. Models are assessed on their ability to identify clinicallyTarget Identification - Cardiovascular diseaseTargetBench · Evaluation of AI models on target identification for Cardiovascular disease. Models are assessed on their ability to ideTarget Identification - Endocrine and metabolic diseasesTargetBench · Evaluation of AI models on target identification for Endocrine and metabolic diseases. Models are assessed on their abilTarget Identification - Fibrotic diseaseTargetBench · Evaluation of AI models on target identification for Fibrotic disease. Models are assessed on their ability to identify Target Identification - Inflammation and ImmunologyTargetBench · Evaluation of AI models on target identification for Inflammation and Immunology. Models are assessed on their ability tTarget Identification - Mental or behavioural disorderTargetBench · Evaluation of AI models on target identification for Mental or behavioural disorder. Models are assessed on their abilitTarget Identification - Neurologic diseasesTargetBench · Evaluation of AI models on target identification for Neurologic diseases. Models are assessed on their ability to identiTarget Identification - OphthalmologyTargetBench · Evaluation of AI models on target identification for Ophthalmology. Models are assessed on their ability to identify cliTarget Identification - Other Diseases - Multi-CausesTargetBench · Evaluation of AI models on target identification for Other Diseases - Multi-Causes. Models are assessed on their abilityTarget Identification - Reproductiveness, Pregnancy and childbirthTargetBench · Evaluation of AI models on target identification for Reproductiveness, Pregnancy and childbirth. Models are assessed on Aging PredictionLongevity Benchmark · General aging-related prediction task.GTEx ExpressionLongevity Benchmark · Gene expression level prediction from GTEx data.GTEx GenerativeLongevity Benchmark · Generative task for GTEx gene expression patterns.GTEx Pairwise BinaryLongevity Benchmark · Binary pairwise comparison of gene expression from GTEx.GTEx Pairwise TernaryLongevity Benchmark · Ternary pairwise comparison of GTEx gene expression.Longevity Synergy (Full)Longevity Benchmark · Predicting synergistic effects of longevity interventions (full context).Longevity Synergy (Minimal)Longevity Benchmark · Predicting synergistic effects of longevity interventions (minimal context).Methylation Age ChoiceLongevity Benchmark · Multiple choice task for methylation-based age prediction.Methylation Age PairwiseLongevity Benchmark · Comparing biological age from DNA methylation patterns.Methylation Age RegressionLongevity Benchmark · Regression task for exact biological age from methylation.NHANES Mortality ClassificationLongevity Benchmark · Binary classification of mortality risk using NHANES health survey data.NHANES Pairwise ComparisonLongevity Benchmark · Pairwise comparison of individuals based on health indicators.NHANES Time-to-EventLongevity Benchmark · Multi-class time-to-event prediction from NHANES data.NHANES TTE RegressionLongevity Benchmark · Regression task for exact time-to-event prediction.Olink GenerativeLongevity Benchmark · Generative task for Olink protein biomarkers.Olink Pairwise ComparisonLongevity Benchmark · Pairwise comparison using Olink protein biomarkers.Olink Protein ClassificationLongevity Benchmark · Classifying samples based on Olink proteomics markers.Synergy RegressionLongevity Benchmark · Regression task for quantifying synergy effects.TCGA Survival PredictionLongevity Benchmark · Predicting cancer patient survival outcomes from TCGA genomic data.
Chemical Synthesis (2)
Single-step retrosynthesis, URSA-expert-2026 datasetISM Benchmarks: Retrosynthesis · Evaluates single-step retrosynthesis task performance on URSA-expert-2026 dataset across various ChemCensor metrics.Single-step retrosynthesis, USPTO-50k-test sample datasetISM Benchmarks: Retrosynthesis · Evaluates single-step retrosynthesis task performance on representative 10% sample from USPTO-50k-test dataset across va
Clinical Trials (25)
All Phases - 2025Q1ClinBench Quarterly · Clinical trial outcome prediction (All Phases) for trials with results submitted in 2025Q1. N=356 samples.All Phases - 2025Q2ClinBench Quarterly · Clinical trial outcome prediction (All Phases) for trials with results submitted in 2025Q2. N=454 samples.All Phases - 2025Q3ClinBench Quarterly · Clinical trial outcome prediction (All Phases) for trials with results submitted in 2025Q3. N=297 samples.All Phases - 2025Q4ClinBench Quarterly · Clinical trial outcome prediction (All Phases) for trials with results submitted in 2025Q4. N=293 samples.All Phases - 2026Q1ClinBench Quarterly · Clinical trial outcome prediction (All Phases) for trials with results submitted in 2026Q1. N=87 samples.Phase 1 - 2025Q1ClinBench Quarterly · Clinical trial outcome prediction (Phase 1) for trials with results submitted in 2025Q1. N=56 samples.Phase 1 - 2025Q2ClinBench Quarterly · Clinical trial outcome prediction (Phase 1) for trials with results submitted in 2025Q2. N=127 samples.Phase 1 - 2025Q3ClinBench Quarterly · Clinical trial outcome prediction (Phase 1) for trials with results submitted in 2025Q3. N=30 samples.Phase 1 - 2025Q4ClinBench Quarterly · Clinical trial outcome prediction (Phase 1) for trials with results submitted in 2025Q4. N=23 samples.Phase 1 - 2026Q1ClinBench Quarterly · Clinical trial outcome prediction (Phase 1) for trials with results submitted in 2026Q1. N=3 samples.Phase 2 - 2025Q1ClinBench Quarterly · Clinical trial outcome prediction (Phase 2) for trials with results submitted in 2025Q1. N=146 samples.Phase 2 - 2025Q2ClinBench Quarterly · Clinical trial outcome prediction (Phase 2) for trials with results submitted in 2025Q2. N=163 samples.Phase 2 - 2025Q3ClinBench Quarterly · Clinical trial outcome prediction (Phase 2) for trials with results submitted in 2025Q3. N=145 samples.Phase 2 - 2025Q4ClinBench Quarterly · Clinical trial outcome prediction (Phase 2) for trials with results submitted in 2025Q4. N=114 samples.Phase 2 - 2026Q1ClinBench Quarterly · Clinical trial outcome prediction (Phase 2) for trials with results submitted in 2026Q1. N=40 samples.Phase 3 - 2025Q1ClinBench Quarterly · Clinical trial outcome prediction (Phase 3) for trials with results submitted in 2025Q1. N=119 samples.Phase 3 - 2025Q2ClinBench Quarterly · Clinical trial outcome prediction (Phase 3) for trials with results submitted in 2025Q2. N=133 samples.Phase 3 - 2025Q3ClinBench Quarterly · Clinical trial outcome prediction (Phase 3) for trials with results submitted in 2025Q3. N=100 samples.Phase 3 - 2025Q4ClinBench Quarterly · Clinical trial outcome prediction (Phase 3) for trials with results submitted in 2025Q4. N=113 samples.Phase 3 - 2026Q1ClinBench Quarterly · Clinical trial outcome prediction (Phase 3) for trials with results submitted in 2026Q1. N=38 samples.Phase 4 - 2025Q1ClinBench Quarterly · Clinical trial outcome prediction (Phase 4) for trials with results submitted in 2025Q1. N=35 samples.Phase 4 - 2025Q2ClinBench Quarterly · Clinical trial outcome prediction (Phase 4) for trials with results submitted in 2025Q2. N=31 samples.Phase 4 - 2025Q3ClinBench Quarterly · Clinical trial outcome prediction (Phase 4) for trials with results submitted in 2025Q3. N=22 samples.Phase 4 - 2025Q4ClinBench Quarterly · Clinical trial outcome prediction (Phase 4) for trials with results submitted in 2025Q4. N=43 samples.Phase 4 - 2026Q1ClinBench Quarterly · Clinical trial outcome prediction (Phase 4) for trials with results submitted in 2026Q1. N=6 samples.