MODULE 01TARGET / HIT DISCOVERY
Q-Dock
HIT DISCOVERY SOLUTION MODULE
Models target–ligand interactions to identify candidates with credible affinity and binding structures.
Binding prediction · Candidate priority · Experimental validation
QUEST™ / Quantum–AI Drug Discovery
Innovating the critical bottlenecks in drug development.
QUEST™ integrates molecular modelling, experimental evidence, artificial intelligence and quantum-derived representations to support better decisions from hit discovery through preclinical prediction.
Q-Dock
Hit Discovery
Hit Discovery
Supernatural AI
Lead Optimization
Lead Optimization
Virtual CRO
Preclinical Prediction
Preclinical Prediction
01 / Where QUEST™ Intervenes
QUEST™ addresses the decisive bottlenecks in drug development.
QUEST™ is QIC’s integrated Quantum–AI drug-discovery solution. Its three specialized modules support the development lifecycle from identifying experimentally credible hits to optimizing candidates and assessing preclinical ADMET and PK risk.
Drug-development lifecycleThis map shows where the major bottlenecks arise and which QUEST™ module addresses each one. Clinical development is shown only as the downstream stage and is not a current QUEST™ module.
01TargetModel
02Hit DiscoveryDiscover
03Hit-to-LeadPrioritize
04Lead OptimizationOptimize
05PreclinicalPredict
06ClinicalDevelop
BOTTLENECK 01
Uncertainty in selecting viable hits
Conventional AI-based virtual screening assigns predictive scores from learned data patterns. A high score, however, does not necessarily indicate actual binding affinity, selectivity or a valid binding pose, and false-positive candidates may fail in experimental validation.
BOTTLENECK 02
Simultaneous optimization of competing drug properties
Strong binding affinity alone does not make a viable drug candidate. Affinity, selectivity, permeability, solubility, metabolic stability and toxicity must be optimized together, even though improving one property may compromise another.
BOTTLENECK 03
Late identification of ADMET and PK liabilities
ADMET and pharmacokinetic issues are often identified only after substantial time and cost have been committed to preclinical studies, potentially leading to candidate termination and development delays.
QUEST™ Solution ModulesONE INTEGRATED SOLUTION · THREE SPECIALIZED MODULES
MODULE 02HIT-TO-LEAD / LEAD OPTIMIZATION
Supernatural AI
SCREENING & LEAD OPTIMIZATION MODULE
Identifies activity-driving fragments and designs candidates with a balanced drug-property profile.
Fragment analysis · Candidate design · Multi-property optimization
MODULE 03PRECLINICAL
Virtual CRO
PRECLINICAL PREDICTION SOLUTION MODULE
Predicts ADMET, PK and exposure profiles to support earlier candidate decisions and molecular optimization.
ADMET/PK · Drug exposure · Molecular optimization
Cross-cutting challenge
Limits of data-only AI. Sparse or biased datasets, unfamiliar chemical space and limited representation of molecular physics can reduce prediction reliability across the lifecycle.
QUEST™ MODULE01
HIT DISCOVERY SOLUTION MODULE
Q-Dock
Q-Dock models both the target binding environment and candidate ligands, then compares their three-dimensional and electronic interaction patterns to predict binding affinity and the most plausible binding pose.
CANDIDATE ANALYSIS WORKFLOWCANDIDATE ANALYSIS
STEP 01
Target binding-site analysis
Define the binding pocket and its spatial and electronic interaction environment.
→
STEP 02
Candidate structure & property analysis
Generate plausible ligand geometries and comparable interaction patterns.
→
STEP 03
Target–candidate interaction prediction
Rank candidates and predict affinity and binding poses.
MODEL TRAINING & DATA FOUNDATION
Chemical DB + Quantum Modeling
Build learning representations from molecular and interaction data
→
3D Image Match ML Model
Train the pattern-recognition and scoring model used in candidate matching
KEY ANALYSIS RESULTSQ-Dock Candidate Assessment
01Candidate priority
02Binding affinity
03Binding pose
04Candidates for experimental validation
QUEST™ MODULE02
SCREENING & LEAD OPTIMIZATION MODULE
Supernatural AI
Supernatural AI identifies the molecular fragments that drive activity and uses them to iteratively design candidates with improved affinity, selectivity and drug-like properties.
MOLECULAR DESIGN & OPTIMIZATIONIDENTIFY → EVALUATE → DESIGN ↺
PHASE 01
Extract activity-related chemical signals
Integrate fragment libraries, assay results, virtual fragmentation and DEL screening data to identify credible activity signals.
Candidate fragment seeds
Assay and DEL evidence
Noise filtering
PHASE 02
Assess binding contribution and developability
Apply quantum and 3D alignment with target analysis to assess binding-site compatibility and developability.
Key binding fragments
Natural / unnatural amino acids
Selectivity · toxicity · solubility
PHASE 03
Design candidates with a balanced profile
Recombine selected fragments and iteratively improve affinity, permeability and other drug-like properties.
Fragment connection
Peptide-to-small-compound design
Affinity · permeability optimization
FEEDBACK LOOPEach designed candidate is reassessed and refined until the target balance of activity and drug-like properties is achieved.
Key Analysis Results
- Validated activity signals
- Key binding fragments
- Structure–activity insights
Candidate Design Results
- Designed candidate structures
- Candidate and chemical-series priority
- Multi-property profile
- Recommended series for experimental validation
QUEST™ MODULE03
PRECLINICAL PREDICTION SOLUTION MODULE
Virtual CRO
Virtual CRO translates molecular structure into ADMET and PK predictions, PK/PBPK analysis and concentration–time profiles, enabling earlier comparison and optimization of preclinical candidates.
PRECLINICAL ADMET & PK ASSESSMENT WORKFLOW6 STAGES
STAGE 01
Drug structure input
Molecular structure or SMILES
STAGE 02
Molecular representation
Quantum-derived and AI-ready molecular features
STAGE 03
PK parameter prediction
Permeability, clearance and related PK parameters
STAGE 04
PK/PBPK model analysis
Compartmental or physiologically based modelling
STAGE 05
Concentration–time curve
Predicted drug exposure in blood and tissues
STAGE 06
Molecular optimization
Candidate redesign toward the target PK profile
Development Decision Support
Compare candidate exposure profilesDetect ADMET and PK liabilities earlierPrioritize follow-up experimentsGuide molecular optimization
Virtual CRO in the NAMs transition
Regulatory science is expanding the use of computational and other New Approach Methodologies to support more human-relevant evidence and the responsible reduction, refinement and replacement of animal testing. Virtual CRO is positioned within a broader weight-of-evidence strategy—not as a standalone replacement for required studies.
QUEST™ IN ACTIONREAL-WORLD DRUG DEVELOPMENT PROGRAM 01
ORAL PEPTIDE DRUG DISCOVERY
QUEST™ advances a proprietary α4β7 program from candidate discovery to oral drug development.
QIC applies QUEST™ to discover and optimize its proprietary α4β7-targeted peptide candidate, including binding affinity, target selectivity, drug-like properties, PK/PD and toxicity prediction. D&D Pharmatech applies and optimizes its ORALINK™ platform and conducts experimental efficacy evaluation to advance the asset toward an orally deliverable drug candidate.
QIC · QUEST™ APPLICATIONONE ASSET · THREE SPECIALIZED MODULES
Q-Dock
Models the α4β7 binding environment and evaluates target–candidate interactions.
- Binding-site analysis
- Affinity and binding-pose prediction
Supernatural AI
Designs and optimizes peptide candidates for activity, selectivity and developability.
- Candidate discovery and design
- Multi-property optimization
Virtual CRO
Predicts PK/PD and toxicity to identify liabilities and guide candidate refinement.
- PK/PD and exposure prediction
- Toxicity and safety assessment
QUEST™
Connect platform capability with real drug-development programs.
Explore QIC’s proprietary pipeline or discuss a discovery, optimization or preclinical prediction program.
View Pipeline → Discuss a Program →