Pharmacophore Modeling & Screening
Ligand-Based Features 
Structure-Based Constraints 
AI-Enhanced Screening We deliver AI-enhanced pharmacophore models --- validated by molecular docking and MD simulation --- to enable scaffold hopping and focused library design with higher hit rates than fingerprint methods.
Why Pharmacophore Modeling Is the Critical Foundation
Structure-based hit identification without pharmacophore guidance is optimization without a hypothesis. Seed-stage biotechs need to prioritize scaffolds that preserve critical interactions; pharma teams require pharmacophore constraints that survive docking and FEP validation. Our platform generates ligand-based pharmacophores from active compounds, structure-based pharmacophores from co-crystal structures, and hybrid models that integrate both --- then validates them through ensemble docking and MD trajectory analysis to ensure that every feature sphere corresponds to a physically meaningful interaction.
What Sets the Platform Apart
Ligand + Structure Hybrid
We derive pharmacophore features from both active ligands (2D/3D QSAR) and protein binding sites (structure-based), then merge them into consensus models that capture the full interaction landscape.
AI-Enhanced Feature Detection
ML classifiers trained on 500+ co-crystal structures automatically detect pharmacophore features from electron density and MD trajectories, reducing manual annotation bias.
Scaffold-Hopping Ready
Every pharmacophore model includes feature tolerance spheres and exclusion volumes, enabling scaffold hopping and bioisostere replacement without losing binding affinity.
Technology Suite
Ligand-Based Pharmacophore Generation
Automated identification of shared pharmacophore features across active ligand sets
Key Features:
- Common-Feature Extraction — Automated identification of shared pharmacophore features (H-bond donors, acceptors, hydrophobic centroids, aromatic rings) across active ligand sets.
- 3D-QSAR Integration — Alignment-free and alignment-dependent pharmacophore models combined with CoMFA/CoMSIA for quantitative activity prediction.
- Ensemble Pharmacophore — Multiple bioactive conformation sampling to capture ligand flexibility and induced-fit contributions.
Ideal For: Programs with multiple known actives but no target structure; scaffold hopping campaigns; natural product target deconvolution.
What We Offer:
From a set of active compounds, we generate a validated pharmacophore hypothesis with exclusion volumes, ready for 3D database screening and library design.
Structure-Based Pharmacophore Modeling
Extraction of pharmacophore features directly from binding pocket geometry
Key Features:
- Pocket-Derived Feature Mapping — Extraction of pharmacophore features directly from binding pocket geometry using AlphaFold or experimental structures.
- Dynamic Pharmacophore — MD-derived ensemble pharmacophores capturing transient pockets and allosteric sites invisible in static models.
- Selectivity Pharmacophore — Comparative feature mapping against off-target homologs to design selective ligands.
Ideal For: Structure-enabled programs requiring rational optimization; selectivity-driven kinase or protease inhibitor design.
What We Offer:
A structure-based pharmacophore with validated feature constraints derived from co-crystal waters, metal ions, and protein side chains. Directly compatible with virtual screening and de novo design pipelines.
Scaffold Hopping Analysis
Systematic substitution of core scaffolds while preserving pharmacophore geometry
Key Features:
- Bioisostere Replacement — Systematic substitution of core scaffolds while preserving pharmacophore geometry and physicochemical properties.
- Shape & Electrostatic Matching — ROCS/EON-style 3D shape and field-point alignment to identify topologically distinct but functionally equivalent chemotypes.
- Patent Escape Routing — Real-time novelty scoring to ensure hop candidates occupy unclaimed chemical space.
Ideal For: Fast-follower programs avoiding competitor IP; lead series with poor ADMET properties requiring core replacement; fragment-to-lead expansion.
What We Offer:
A ranked list of scaffold-hop candidates with 3D alignment scores, predicted affinity retention, and synthetic accessibility. Each candidate is pre-screened for patent novelty and ADMET liabilities.
3D-Pharmacophore Search
GPU-accelerated 3D pharmacophore matching against billion-compound libraries
Key Features:
- Database Screening Speed — GPU-accelerated 3D pharmacophore matching against billion-compound libraries (Enamine REAL, Mcule, ZINC).
- Fuzzy Matching — Tolerance-adjusted feature matching to account for conformational uncertainty and partial feature satisfaction.
- Hit Enrichment Reporting — Statistical validation of pharmacophore enrichment against decoy sets with ROC AUC and EF metrics.
Ideal For: Ultra-large library screening; DEL follow-up campaigns; rapid idea validation before synthesis investment.
What We Offer:
Screening results within hours against commercial and proprietary databases, with annotated hit lists ready for docking and MD validation.
Platform Instrumentation
Core Instruments
| Instrument / Software | Capability |
|---|---|
| NVIDIA DGX A100 | AI pharmacophore feature detection and large-scale 3D screening |
| Schrödinger Phase / Catalyst | Ligand-based pharmacophore modeling and database search |
| RDKit / OpenEye | Pharmacophore fingerprint generation and similarity scoring |
| GROMACS/AMBER HPC | MD-based dynamic pharmacophore refinement and validation |
Standardized Workflow
Project Workflow
A milestone-driven execution system from active compounds to enriched hit library.
01 Target Review
- Active compound collection
- SAR data review
- Target structure assessment (if available)
- Deliverable: Target assessment report + feature proposal
02 Feature Extraction
- Ligand-based feature extraction
- Structure-based pocket mapping
- MD trajectory feature sampling
- Deliverable: Feature inventory + confidence scores
03 Model Building
- Pharmacophore model assembly
- Tolerance sphere optimization
- Exclusion volume definition
- Deliverable: 3D pharmacophore model + query file
04 Screening
- Database screening (corporate / commercial)
- Scaffold hopping analysis
- Hit clustering and diversity analysis
- Deliverable: Enriched hit list + scaffold report
05 Validation
- Docking validation of top hits
- MD interaction confirmation (optional)
- Deliverable: Validated hits + binding mode report + final report
Sample Requirements
- Active compounds: SMILES, SDF, or compound IDs with activity data (IC50, Ki, EC50)
- Inactive compounds (optional): For feature discrimination and model refinement
- Target structure (if available): PDB file or AlphaFold model for structure-based pharmacophore
- Project scope: Desired scaffold diversity, patent constraints, and ADMET optimization goals
- Screening library preference: Corporate collection, commercial vendor (Enamine, ChemDiv, ZINC), or focused subset
Standard Deliverables
- 3D pharmacophore model with feature spheres, tolerance radii, and exclusion volumes (PDB / MOE / Catalyst format)
- Enriched hit list from database screening with pharmacophore match scores and property profiles
- Scaffold hopping report with novel chemotypes preserving critical features
- Bioisostere replacement suggestions with predicted ADMET impact
- Docking validation data for top-ranked hits (poses and interaction fingerprints)
- MD validation report (if selected): dynamic feature stability and interaction persistence
- Final technical report with SAR recommendations and synthesis priorities
Frequently Asked Questions
Case Study
Case Study: Consensus Pharmacophore Strategy Identifies Novel SARS-CoV-2 Mpro Inhibitors from Large Chemical Libraries
Goal: Validate a pharmacophore-based virtual screening pipeline that identifies novel non-covalent Mpro inhibitors from multi-million compound libraries, establishing precedent for scaffold hopping against pandemic viral targets.
Key Data:
- Pharmacophore model: Generated from 16 SARS-CoV Mpro inhibitor classes + pan-coronavirus inhibitor X77 (PDB: 6w63), validated through Fischer validation and cross-docking.
- Library scale: Screened 213.5M compounds across 9 Pharmit molecular libraries; 711,102 hits matched the 17-feature consensus pharmacophore.
- Hit filtering: Docking + MM-GBSA + ADMET narrowed hits to 257 promising inhibitors; MD simulations confirmed 3 effective candidate inhibitors (ECIs) maintaining stable Mpro binding.
- Binding mechanism: van der Waals interactions dominate affinity over electrostatic forces; competitive H-bonds and electrostatic desolvation penalties are unfavorable --- guiding future optimization strategy.
Why it matters: For drug developers facing emerging viral targets without established chemotypes, this study demonstrates that pharmacophore-guided screening of ultra-large libraries compresses hit identification timelines from years to months. By combining structure-based pharmacophore constraints with MD-validated binding stability, teams can prioritize synthesis candidates with pre-validated interaction patterns --- directly supporting antiviral programs where traditional HTS is impractical.
Figure 1. Similarity matrix of the selected compounds. The compounds in the reference set were included for comparison. (Ruiz-Moreno AJ.; et al. 2024)
Reference
Ruiz-Moreno AJ, et al. Consensus Pharmacophore Strategy For Identifying Novel SARS-Cov-2 Mpro Inhibitors from Large Chemical Libraries. J Chem Inf Model. 2024 Mar 25;64(6):1984-1995.
Our technical team responds within 24 hours. All inquiries protected under NDA.
