Now Available

Discover Drug Candidates Faster

SSI swarm intelligence explores multi-property molecular landscapes on GPU, scoring compounds against 15,000+ validated ChEMBL targets. Virtual screening campaigns that would take weeks run in hours.

Start Free Trial API Documentation

Discovery Pipeline

From target selection to lead candidates in a single platform.

1

Select Targets

Choose from 15,000+ validated ChEMBL target models or upload your own binding assay data.

2

Explore Landscape

SSI swarm agents map the multi-property molecular landscape on GPU, finding activity cliffs and sweet spots.

3

Screen & Score

GPU-accelerated virtual screening scores up to millions of compounds against your targets with ADMET filtering.

4

Get Candidates

Receive ranked hit lists with Pareto-optimal candidates, activity predictions, and ADMET profiles. Export as CSV or SDF.

Platform Capabilities

Built on the same distributed swarm engine that powers Mesh Optimizer and Materials Discovery.

Multi-Property Landscape Optimization

SSI swarm agents explore molecular space across multiple objectives simultaneously — activity, selectivity, ADMET properties, synthetic accessibility — finding Pareto-optimal regions that single-objective methods miss.

  • Multi-objective optimization
  • Activity cliff detection
  • Sweet spot identification
  • Pareto front extraction

15,000+ Validated Targets

Pre-built target models from the ChEMBL database, each validated against known actives and decoys. Select targets by name, UniProt ID, or disease area. Upload custom binding data for proprietary targets.

  • ChEMBL 34 target models
  • Disease-area browsing
  • Custom target upload
  • Multi-target campaigns

GPU-Accelerated Virtual Screening

Score compound libraries at scale with GPU-accelerated molecular docking and scoring functions. Process up to millions of compounds per campaign with hardware-aware scheduling.

  • Up to 1M+ compounds per run
  • Molecular docking scores
  • Fingerprint similarity
  • Shape-based screening

ADMET Property Prediction

Filter candidates early with integrated ADMET property prediction. Absorption, distribution, metabolism, excretion, and toxicity predictions for every scored compound.

  • Lipinski / Veber rule filtering
  • CYP inhibition prediction
  • hERG liability scoring
  • Solubility & permeability

De Novo Molecule Generation

Generate novel molecular structures optimized for your target profile. SSI agents explore chemical space beyond known compound libraries to find genuinely new scaffolds.

  • Scaffold hopping
  • Property-guided generation
  • Synthetic accessibility scoring
  • Novelty vs. similarity controls

Campaign Orchestration

Run multi-target screening campaigns with automatic hit triaging, target deconvolution, and selectivity profiling. Track campaigns from initial screen through lead optimization.

  • Multi-target hit triaging
  • Selectivity profiling
  • Campaign history & comparison
  • REST API for pipeline integration

Simple API

Submit a discovery job in one request. Poll or use webhooks for results.

Submit a Screening Job

curl -X POST https://api.pharma.slentosystems.com/jobs/submit \
  -H "X-API-Key: pd-pro-your-key" \
  -H "Content-Type: application/json" \
  -d '{
    "target_id": "CHEMBL220",
    "discovery_mode": "standard",
    "library": "zinc_druglike_subset",
    "max_compounds": 100000,
    "admet_filter": true,
    "callback_url": "https://your-app.com/webhook"
  }'

Get Results

{
  "job_id": "pd-a1b2c3d4-...",
  "target": "CHEMBL220 (GPCR)",
  "compounds_screened": 100000,
  "hits": 347,
  "top_candidate": {
    "smiles": "CC(=O)Nc1ccc(O)cc1",
    "score": 0.94,
    "admet": {"lipinski": true, "hERG": "low"},
    "dock_score": -9.2
  },
  "pareto_front": [ ... ],
  "export_url": "/results/pd-a1b2c3d4.sdf"
}

Two Ways to Use

SaaS (Managed)

Submit screening jobs via API. We run the swarm on our GPU cluster and return scored, ranked results. No infrastructure needed on your end.

  • Zero infrastructure to manage
  • Results in hours, not weeks
  • REST API + webhook callbacks
  • Pay per job or subscribe monthly
View Pricing

Standalone (Self-Hosted)

Run the pharma discovery engine on your own GPU hardware. Full control over your compound data and target models. Air-gapped compatible. Download available after purchasing a license.

  • Your data stays on your machines
  • AMD RDNA3 + NVIDIA GPU support
  • Integrate with RDKit, OpenBabel, AutoDock
  • BSL 1.1 license (Apache 2.0 after 4 years)
Get a License

Ready to Discover?

Free trial: 2 exploration runs, 5,000 scoring credits. No credit card required.

Start Free Trial Talk to Sales