Home Structural BiologyVirtual Screening

Drug Discovery Services

In Silico Virtual Screening
Accelerate Your Hit Discovery

Virtual screening uses computer-aided drug design (CADD) to select promising compounds from large molecular libraries for further experimental evaluation — combining structure-based docking, ligand-based pharmacophore modeling, and deep learning to achieve hit rates of 5–20%, far exceeding traditional high-throughput screening.

Structure-BasedLigand-BasedHybrid StrategyQSAR ModelingADMET Prediction

Request a Consultation

5–20%

Typical Reported Hit Rate

10M+

Public Library Compounds

Complementary Strategies

CADD

Comprehensive Platform

Deep Learning Integration

Service Overview

Identify Novel Leads from Millions of Molecules — In Silico

Virtual screening is a computational approach that selects promising compounds from large chemical databases by applying drug design theory, physics-based docking, and machine learning. ICDMO's comprehensive CADD platform enables customized in silico virtual screening services — from rapid pharmacophore filtering to multi-million compound docking campaigns — dramatically shortening the time and cost of early drug discovery.

By combining the results of virtual screening with experimental assays, the experimental effort becomes more targeted, the hit rate increases substantially, and the research cycle is shortened. Our positive hit rate typically reaches 5–20%, far exceeding conventional high-throughput screening approaches.

Suitable For

Drug targets with known 3D structure (crystal, cryo-EM, NMR)

Targets where only homologous structures are available

Projects with known active compounds for QSAR / pharmacophore

PPI (protein–protein interaction) hot-spot pocket inhibition

Fragment-to-lead expansion and scaffold hopping campaigns

Focused library design for a specific target class (kinase, GPCR, protease)

Repositioning studies: screening approved drug libraries against new targets

Projects requiring rapid, cost-effective hit identification before HTS

Molecular Docking · Binding Site Analysis

Strategy 1

3D Docking

Target Structure-Based Virtual Screening

Using the known 3D structure of the target, we analyze the binding site and predict the binding conformation of target–ligand complexes by molecular docking. Scoring functions evaluate protein–ligand binding modes, and pharmacophore modeling with MD simulations refine the selection of high-affinity candidates.

QSAR · Pharmacophore · Shape Similarity

Strategy 2

QSAR

Ligand-Based Virtual Screening

When a 3D target structure is unavailable, we analyze the structures, physicochemical properties, and structure–activity relationships of compounds with known activities to establish QSAR and pharmacophore models. This approach has the advantages of fast speed and versatility — not limited by the availability of a target structure.

Structure-Based + Ligand-Based Combined

Strategy 3

Hybrid

Combined Strategy for Maximum Hit Rate

ICDMO integrates target structure information and ligand similarity in a hierarchical, parallel, or hybrid manner depending on project specifics. This combined approach maximizes coverage of chemical space, improves enrichment factors, and is particularly powerful for challenging targets such as PPIs and allosteric sites.

Service Workflow

From Target Information to Confirmed Hit Compounds

A systematic five-stage quality-controlled pipeline covering target investigation through final hit delivery and scientific report.

Target Investigation & Binding Site Analysis

Client provides target information: sequence, PDB ID, or homologous structure reference

Binding site identification: SiteMap, fpocket, or cavity detection algorithms

Target protein preparation: protonation, hydrogen addition, energy minimization

Druggability assessment and allosteric pocket characterization for novel targets

Compound Database Preparation & ADMET Pre-Filtering

Access to commercial and public libraries: ZINC, ChemBridge, Enamine, Maybridge — licence terms confirmed per project

Lipinski Rule-of-Five and ADMET property pre-filtering to reduce inactive compounds

Structure standardization: salt removal, tautomer enumeration, 3D conformer generation

Diversity-based subset selection; focused library curation around target class

Pharmacophore Modeling & QSAR Analysis

Structure-based pharmacophore generation from protein-ligand complex interactions

Ligand-based pharmacophore from known actives when no 3D structure is available

QSAR model development: physicochemical descriptors, 2D/3D fingerprints, ML models

Shape and electrostatic similarity screening (Phase, MOE, Pharmit) for rapid first-pass filtering

Molecular Docking & Scoring

Hierarchical docking: HTVS → Standard Precision (SP) → Extra Precision (XP)

Several independent docking methods run in parallel, with consensus scoring used to raise confidence and disagreement reported

Binding mode analysis: hydrogen bond, hydrophobic, halogen bond, pi-stacking interactions

Refinement with molecular dynamics (MD) and MM-GBSA free energy re-scoring

Hit Identification, Manual Selection & Delivery

Expert medicinal chemist review of top-scored compounds for synthetic accessibility

Cluster analysis to maximize structural diversity in final hit list

Predicted ADMET profiling: logP, solubility, CYP inhibition, hERG, BBB penetration

Delivery of shortlisted active compounds with full docking poses and scoring report

Why Choose ICDMO

Service Advantages

Advanced Target Structure Analysis

ICDMO uses cutting-edge binding site detection and structural analysis tools to characterize cryptic pockets and allosteric sites overlooked by conventional approaches, maximizing the chances of finding genuine binders.

Extensive Virtual Compound Database

Access to tens of millions of commercially available compounds spanning ZINC, Enamine REAL, ChemBridge, and our own curated focused libraries — covering a broad chemical space for any target class.

Comprehensive CADD Platform

Our integrated CADD platform combines molecular docking, pharmacophore modeling, QSAR, MD simulation, and free energy methods in a single quality-controlled pipeline, ensuring reproducible and scientifically rigorous outcomes.

Deep Learning Model Integration

ICDMO integrates graph neural networks and transformer-based scoring models alongside traditional docking, significantly improving hit rate and reducing false positives compared to classical virtual screening alone.

Flexible & Cost-Effective Solutions

We tailor every project to your specific needs and budget — whether a rapid pharmacophore filter on a focused library or a full multi-million compound docking campaign — delivering maximum value for your discovery investment.

Confidential Project Management

ICDMO adopts a strict private management and confidentiality policy for all project information and data. NDAs are standard, and your proprietary target data and hit lists remain fully protected throughout the engagement.

Case Studies

Representative Virtual Screening Projects

EGFR T790M Mutation-Selective Inhibitor Discovery

Kinase Target

Hit Rate: 18%

EGFR T790M Mutation-Selective Inhibitor Discovery

▸Target: EGFR T790M / C797S double-mutant kinase (osimertinib-resistant)

▸Strategy: Structure-based virtual screening (SBVS) with covalent docking

▸Library: 4.2 million commercially available, purchasable compounds

▸Docking: covalent docking with precision scoring, followed by end-point free energy re-scoring; 312 compounds selected

▸Outcome: 18% confirmed hit rate in biochemical IC₅₀ assay (n = 312 compounds)

Structure-based virtual screening against the ATP-binding pocket of the EGFR T790M/C797S double mutant identified 312 covalent docking candidates from a 4.2M compound library. Biochemical testing confirmed 56 hits (IC₅₀ < 10 µM), with two scaffolds demonstrating > 100-fold selectivity over wild-type EGFR — providing novel starting points for next-generation NSCLC inhibitor development.

MDM2-p53 Protein–Protein Interaction Disruptor Identification

PPI Target

Hit Rate: 12%

MDM2-p53 Protein–Protein Interaction Disruptor Identification

▸Target: MDM2 p53-binding cleft (Phe19/Trp23/Leu26 hot-spot pocket)

▸Strategy: Hybrid SBVS + pharmacophore filtering for PPI hot-spot pocket

▸Library: 1.8 million drug-like compounds + 200K natural product derivatives

▸Method: pharmacophore pre-filtering, staged docking from standard to extra precision, then end-point free energy refinement

▸Outcome: 12% confirmed hit rate; 3 novel scaffolds with Kd < 5 µM (SPR)

A hybrid virtual screening workflow combining pharmacophore-based pre-filtering with high-precision molecular docking successfully identified novel MDM2 inhibitors targeting the p53 hot-spot binding cleft — a challenging shallow PPI interface. SPR binding assays confirmed 22 hits from 185 experimentally tested compounds, with three scaffolds showing low-micromolar affinity and selective p53 pathway re-activation in HDM2-overexpressing cell lines.

Services Included

Service Item	Description	Turnaround
Target Investigation & Preparation	Binding site detection, receptor grid generation, homology modeling if no crystal structure	1–3 days
Structure-Based Virtual Screening	Hierarchical docking (HTVS → SP → XP); multi-engine consensus scoring; binding mode analysis	1–2 weeks
Ligand-Based Virtual Screening	Pharmacophore screening, shape similarity, QSAR prediction for targets without 3D structure	1–2 weeks
Hybrid Virtual Screening	Sequential or parallel combination of SBVS and LBVS filters; customized for each project	2–3 weeks
Pharmacophore Modeling	Structure-based or ligand-based 3D pharmacophore generation (Phase, MOE); validation on known actives	3–5 days
QSAR Modeling	ML-based QSAR models (RF, SVM, GNN) trained on activity data; activity prediction for screened compounds	1–2 weeks
MM-GBSA Re-scoring	Binding free energy estimation for top-ranked docked poses; improves hit-rate over docking score alone	3–7 days
ADMET Property Prediction	In silico ADMET profiling: logP, solubility, CYP inhibition, hERG, BBB, oral bioavailability	2–3 days
Focused Library Design	Custom compound library curation or design around specific pharmacophore or scaffold for follow-on screening	Custom
Hit List Delivery & Report	SDF/CSV hit list, docking poses, interaction diagrams, ADMET table, full written scientific report	2–3 days

Frequently Asked Questions

Note: Timelines are estimates for standard projects. Custom focused library design, large-scale multi-million compound campaigns, or projects requiring homology modeling and MD-based pocket sampling may require additional discussion. Contact our team for a project-specific timeline and proposal.

Free Feasibility Consultation

Describe your target and available information. Our scientists will recommend the optimal screening strategy and provide a project proposal — at no charge before commitment.

service@icdmo.com +1-631-626-9181

Standard Deliverables

✓Hit compound list (SDF / CSV with structures)

✓3D docking pose files (Maestro / MOL2 / PDB)

✓2D protein–ligand interaction diagrams

✓Predicted ADMET property table

✓MM-GBSA binding energy ranking

✓Pharmacophore hypothesis figure

✓Full written scientific report

✓Vendor ordering information for confirmed hits

Quick Inquiry

Related Drug Discovery Services

X-Ray Crystallography Cryo-Electron Microscopy Structure-Based Drug Design Molecular Dynamics Simulation Structural Biology Overview

What We Need From You

▸Target structure, experimental or predicted, and the binding site to be screened

▸The compound library, or a request that we select and prepare a suitable commercial or on-demand set

▸Known actives and inactives, which allow the protocol to be validated before the real screen runs

▸Selection constraints: purchasability, budget per compound, supplier preference, chemistry to avoid

▸Your assay format, throughput and detection limits, which determine how large a shortlist is useful

How the Work Is Performed

▸Receptor preparation with protonation, tautomer and structural water decisions documented

▸Ligand preparation covering ionisation at assay pH, tautomers, stereochemistry and conformers

▸Retrospective validation against known actives to establish enrichment before screening begins

▸Staged docking from fast triage through precision scoring, with pharmacophore or interaction constraints

▸Rescoring of survivors, scaffold clustering and individual visual inspection of every shortlisted pose

▸Novelty and property analysis of the final selection, with supplier sourcing where compounds are to be bought

What You Receive

▸Ranked shortlist with scores, poses, predicted properties and purchasing information

▸Interaction analysis per compound, including unfavourable or strained features

▸Scaffold diversity summary showing which chemotypes the selection covers

▸Protocol documentation with filter thresholds, attrition at each stage and enrichment results

▸A realistic statement of expected hit rate and what a negative screen would mean

Interpretation and Limitations

Docking scores rank plausibility rather than potency, and compounds separated by small score differences should be treated as indistinguishable. Hit rates vary widely between targets and cannot be promised; where known actives exist we report the enrichment achieved on them. Screening results are hypotheses for biochemical or biophysical testing, and confirmation remains an experimental question.

Questions We Are Asked Before Scoping

What hit rate should we expect?

Published campaigns commonly report a few percent for well-prepared systems, with a wide spread. We do not promise a rate; we report retrospective enrichment on your own actives when you have them.

Do we need a crystal structure?

No — predicted and homology models are frequently sufficient, but the quality of the binding site region governs the outcome. We assess site quality separately and say whether it will carry a screen.

What if the screen produces nothing?

That is reported as a result, with a diagnosis: site quality, library composition, protocol or target tractability. A null screen that explains itself is more useful than a shortlist nobody can justify.

What Published Work Reports

The results below were published by independent groups and are cited to show what these computational approaches have achieved in the literature. They are not ICDMO project results, and outcomes vary substantially between targets, datasets and programmes. We do not promise comparable results for any particular project.

A 22% confirmed hit rate from docking 138 million compounds

In a landmark 2019 study, ~138 million make-on-demand compounds were docked against the dopamine D4 receptor. Of 549 compounds selected and tested, 122 were confirmed active — a 22% hit rate — including ligands in the sub-nanomolar range. The same work screened ~99 million compounds against AmpC β-lactamase, where 5 of 44 tested compounds were confirmed (11%), one of which was among the most potent non-covalent AmpC inhibitors reported at the time.

Lyu et al., Nature, 2019 ↗

Hit rate rises with rank, which is why library scale matters

The same body of work showed that the probability of a compound being confirmed active increases as its docking rank improves, and that larger libraries push better-scoring molecules into the selected set. This is the practical argument for screening make-on-demand space rather than a physical catalogue — and also the reason scoring accuracy, not library size alone, remains the limiting factor.