From target to clinic — in one conversation
Identify targets from 108,000 genomic associations. Validate against clinical evidence. Screen, optimize, dock, simulate — with compliance checked at every step. Your AI orchestrates the full funnel. You make the decisions that matter.
Every stage, one platform
Start with the right target
108,000 target-disease associations ranked by composite evidence score — genetics, expression, and druggability. Adversarial validation stress-tests each hypothesis against clinical trials, ChEMBL bioactivity, and contradicting literature before you commit compute credits.
NovoMCP
Target Discovery — NSCLC
108K
Target-disease pairs
56
Pharmacogenes
135K
Resistance variants
Ranked Targets
Genetic · Expression · Druggable
0.94
score
Genetic · Druggable
0.89
score
Expression · Druggable
0.81
score
Genetic
0.76
score
Optimize without hand-offs
Scaffold hopping generates structurally novel variants. Property-directed optimization targets specific profiles. Every variant auto-enriched with ADMET predictions and compliance screening — without switching tools or exporting intermediates.
Binding, validated
GPU-accelerated AutoDock scores candidates against any protein target. Strain energy validation catches artifact poses. Contact residues, binding distances, and delta affinities — rendered in the conversation, not a separate viewer.
NovoMCP
Docking Results (1PTH)
PDB ID
1PTH
Resolution
1.8 Å
High Quality
Method
X-Ray
Binding Site
Known (co-crystal)
Binding Affinity (4 molecules)
| # | SMILES | kcal/mol | Δ | Contacts |
|---|---|---|---|---|
| 1 | CC(=O)Oc1ccccc1C(=O)O | -9.4 | — | 4 |
| 2 | CC(=O)Oc1ccc(O)cc1C(=O)O | -8.7 | +0.7 | 5 |
| 3 | CC(=O)Oc1cc(F)ccc1C(=O)O | -7.9 | +1.5 | 3 |
| 4 | c1ccc(NC(=O)C2CC2)cc1 | -6.2 | +3.2 | 2 |
Interactions — Best Binder (−9.4 kcal/mol)
TYR-385
H-bond (2.8 Å)
ARG-120
H-bond (3.1 Å)
VAL-349
Hydrophobic (3.9 Å)
LEU-352
Hydrophobic (4.1 Å)
Dynamics, not snapshots
GROMACS molecular dynamics at production scale — 100 ns trajectories with RMSD, RMSF, and hydrogen bond analysis. Jobs run async on GPU; results stream back with full trajectory plots. Your binding pose, validated across time.
NovoMCP
MD Simulation — COX-2 / Aspirin
Duration
100 ns
Temperature
300 K
Avg RMSD
1.8 Å
Binding ΔG
−8.2
kcal/mol
RMSD Trajectory
Backbone Cα
RMSF (avg)
0.9 Å
H-bonds
3.2 avg
Radius Gyration
22.4 Å
The discovery funnel
Target discovery
Identify drug targets from 108,000 omics-derived target-disease associations. Ranked by composite evidence score with suggested PDB structures.
Target validation
Adversarial stress-test against clinical trials, ChEMBL bioactivity, and literature evidence. Tiered confidence scoring before committing credits.
Literature & seed selection
Semantic search across 14,000 papers, 2,400 patents, and 2.4M ChEMBL compounds. Find starting scaffolds with known activity.
ADMET screening
31 ML models predict absorption, distribution, metabolism, excretion, and toxicity. Filter by safety profile before optimization.
Compliance check
FAVES regulatory screening across DEA, FDA, EPA, CWC, EU REACH, BTWC, Australia, OPCW. Controlled substance detection with scaffold matching.
Lead optimization
Scaffold hopping via RDKit substructure replacement. Property-directed optimization via NVIDIA MolMIM. Every variant compliance-screened.
Docking
GPU-accelerated AutoDock against protein targets. Binding affinities, pose analysis, strain energy validation. 3-10 seconds per molecule.
Molecular dynamics
GROMACS GPU simulation — 100 ns trajectories with equilibration analysis. RMSD convergence, RMSF flexibility, hydrogen bond stability.
Patient stratification
Pharmacogenomic analysis across 56 pharmacogene profiles. CYP metabolizer phenotypes, resistance variants, population-level clinical viability.
Start a discovery pipeline
One conversation. Target to clinic. Sign up and run your first funnel in under five minutes.