Skip to main content
Research platform · Open source

AI-integrated
drug discovery platform

Centralizes target identification, experimental data, compound information, and project workflows across the drug discovery pipeline, with embedded AI modules at each stage.

DeployRead the docs
25+
Organizations
Gates
Foundation Funded
TBDA
Tuberculosis Drug Accelerator Consortium
01Gene Pool
02Target Prioritization
03Screening
04Hit Assessment
05Portfolio & Post-Portfolio
Stage 01

Gene Pool

Target identification from genomic & literature sources

GenomicsLiterature MiningTarget ID
MolecuLogix

The molecular
intelligence layer.

Register compounds, search chemical space, analyze activity, and trace compound evolution from screening to portfolio.

Drug molecule structure
Assay activity
IC₅₀, MIC, and assay results linked to each compound.
Molecular properties
MW, TPSA, LogP, and key descriptors at every stage.
Compound evolution
Track optimization history across analog and series changes.
Structure-based search
Substructure, similarity, and exact match across the registry.
AI Capabilities

AI modules for
drug discovery bottlenecks.

compoundGefitinibisantargetEGFRinhibitor,bioactivityIC50=0.033µMfordiseaseNSCLC.
↕
ZD1839
detected
label: ZD1839
paired ✓
COc1cc2ncnc(Nc3ccc(F)c(Cl)c3)c2cc1OC
DOCUSTORE AI
● ACTIVE

Extracting knowledge from unstructured research documents to accelerate insight discovery.

1/ 4
Drug Discovery Pipeline

Scope of DAIKON

Covers the spectrum from target identification through pre-clinical development in early drug discovery.

OUT OF SCOPE
01
Target Identification
  • Gene pool curation
  • Literature mining
  • Genomic databases
Literature AI · Knowledge graph · DocuStore AI
02
Target Validation
  • Druggability scoring
  • Pathway analysis
  • Genetic evidence
Parsnip · Target scoring AI · DocuStore AI
03
Compound Screening
  • HTS assays
  • Phenotypic screens
  • Nuisance detection
  • ADMET profiling
Nuisance Detection AI · ADMET AI · DocuStore AI
04
Hit Assessment
  • Hit confirmation
  • ADMET profiling
  • Analogs
ADMET AI · DocuStore AI
05
Lead Identification
  • SAR analysis
  • ADMET profiling
  • Series selection
SAR insights · DocuStore AI
06
Lead Optimization
  • IC₅₀ series
  • Selectivity profiling
  • PK/PD modelling
Compound evolution · ADMET AI · DocuStore AI
07
Pre-clinical Test
  • IND filing prep
  • Toxicology
  • Regulatory package
DocuStore AI
08
Preclinical Studies (Animal)
  • Pharmacokinetics
  • Animal models
  • Safety studies
09
Clinical Trials (Human)
  • Phase I · II · III
  • Patient recruitment
  • Endpoint analysis
10
Approval / Launch to Market
  • NDA submission
  • Regulatory review
  • Label approval
Why it matters
01 · Open source
Modular architecture.

Built on a microservices architecture — each module deploys and scales independently. Adaptable to any disease program. Includes target-based & phenotypic discovery stages.

02 · Advisory AI
Human-in-the-loop.

AI suggestions are advisory, not automated. Researchers always approve, reject, or override before anything advances to the next stage.

03 · Provenance
Traceable lineage.

Attribution, notes, milestone dates, and version histories at every stage. Failed experiments and abandoned compounds are preserved — not deleted.

25+Organizations in active use
Currently deployed atTuberculosis Drug Accelerator (TBDA)
Consortium size25+ industry & academic organizations
Program funderGates Foundation
LicenseOpen source · MIT
ArchitectureModular — adaptable to any disease program
Pipeline coverageGene → Target → Screen → Hit → Portfolio → Post-portfolio
Get Started on GitHub

Open source. For academic and research programs.

Start building your
discovery platform today

DAIKON is free, open source, and ready to deploy. Join research programs across academia and industry running their drug discovery portfolios on DAIKON.

Get Started on GitHubRead the Documentation