Start with a target structure, known active ligand, compound library, bioactive molecule, or multi-target question. CD ComputaBio builds a project-specific screening workflow and delivers a ranked, structurally interpreted, and experiment-ready candidate set.
Target, ligand, and library evidence converge on an experiment-ready shortlist.
Match your current evidence to the information needed, the screening program we can build, and the decision-ready outputs you will receive.
A target, ligand, query compound, compound library, or target panel is enough to begin a feasibility discussion.
The workflow can combine structural, ligand, pharmacophore, target, chemistry, and biological evidence.
The final output explains what to test, why it was selected, and what evidence should be generated next.
| Your Starting Point | You Provide | We Build | You Receive |
|---|---|---|---|
| A target structure or reliable model |
|
|
|
| A target but no dependable 3D structure |
|
|
|
| Known active or reference ligands |
|
|
|
| A bioactive compound with an unknown target |
|
|
|
| A proprietary or commercial compound library |
|
|
|
| Multiple targets, mutants, homologs, or anti-targets |
|
|
|
Send the material you already have. One credible starting point and a clear decision are enough for an initial feasibility review. We can identify missing inputs, assess screening readiness, and recommend the most defensible route before large-scale computation begins.
The result is not simply a docking-score list. We connect the raw ranking to structural evidence, chemistry review, candidate-level decisions, and the experiments required to confirm the computational hypothesis.
Traceable source identifiers and primary screening scores.
| Compound ID | Docking Score |
|---|---|
| ZINC000931524028 | −6.574 |
| ZINC000005081146 | −6.570 |
| ZINC000016927727 | −6.570 |
| ZINC000296759835 | −6.555 |
| ZINC000296100325 | −6.529 |
| ZINC000746587264 | −6.520 |
| ZINC000278475905 | −6.509 |
| ZINC000863082599 | −6.395 |
Complete ranking tables can be delivered in CSV or SDF format with preparation states, source-library identifiers, scores, and exclusion history.
Raw and normalized results, identifiers, filters, target annotations, and candidate tiers.
Docked poses, binding-pocket views, pharmacophore maps, and key interaction figures.
Diversity, properties, alerts, availability, pose confidence, and reasons for exclusion.
Primary and backup sets, controls, suggested assays, and the next learning cycle.
Multiple evidence types are reviewed before a candidate is recommended for purchase, synthesis, or experimental testing.
| Priority | Compound ID | Target | Docking Score | MM/GBSA ΔG | Docking LE | Pose / Contacts | Properties | Risk Flags | Recommendation |
|---|---|---|---|---|---|---|---|---|---|
| 01 | VS-104 | C5 | −10.8 | −43.2 | 0.36 | Plausible pose | MW 412; cLogP 2.7 | None detected | Primary Test Set |
| 02 | VS-208 | C5 | −10.1 | −39.6 | 0.40 | Key contacts met | MW 356; cLogP 3.1 | None detected | Primary Test Set |
| 03 | VS-319 | C5 | −9.8 | −38.8 | 0.36 | Plausible pose | MW 391; cLogP 2.4 | None detected | Backup Candidate |
| 04 | VS-426 | C5 | −11.4 | −48.5 | 0.29 | Pose uncertain | MW 521; cLogP 4.8 | High lipophilicity | Manual Review |
| 05 | VS-532 | C5 | −10.6 | −40.7 | 0.31 | Key contacts met | MW 467; cLogP 3.7 | Solubility risk | Manual Review |
Candidate recommendations are based on evidence convergence rather than the most favorable docking score alone. Project-specific fields and thresholds are agreed during scoping.
Structural outputs connect the prepared target model to the predicted binding mode and the interactions used during candidate review.
Scientific interpretation: docking poses are binding hypotheses, MM/GBSA values are computational rescoring estimates rather than experimentally measured free energies, and protein-quality plots assess model readiness rather than compound activity. Final binding and activity claims require experimental confirmation.
The screen is designed around the customer problem and experimental endpoint—not around running one algorithm on every project.
A validated target needs chemical starting points.
A confirmed hit needs nearby chemistry and preliminary SAR coverage.
Known ligands work, but the program needs differentiated chemotypes.
Approved or clinical compounds need evaluation against a new target or disease hypothesis.
A bioactive compound or phenotypic hit lacks a confirmed target.
Shallow, allosteric, PPI, or underexplored pockets need compact starting points.
Related proteins, safety anti-targets, or target mutations must be considered early.
The chemical space is too large for uniform high-cost calculations or experiments.
Combine target-pocket evidence with known-ligand patterns to prioritize diverse, testable chemical starting points.
Evaluate a bioactive compound across a defined target panel and prioritize experimentally testable mechanism hypotheses.
Balance cross-target profiles, chemical diversity, uncertainty, and screening cost in a staged selection funnel.
One method may be sufficient, or several methods can be combined when independent evidence improves the selection decision.
Starting point: A qualified target structure and binding site.
Primary output: Ranked compounds, docking poses, and interaction evidence.
View this screening routeStarting point: Reliable active ligands but no dependable target structure.
Primary output: Analogs, scaffold hops, and model-supported candidates.
View this screening routeStarting point: A bioactive compound with an unknown or incomplete target profile.
Primary output: Ranked targets, off-target flags, and a validation route.
View this screening routeStarting point: A shallow, novel, or challenging pocket.
Primary output: Ligand-efficient fragments and optimization vectors.
View this screening routeStarting point: Known essential interactions or several active chemotypes.
Primary output: Feature-matched, scaffold-diverse candidates.
View this screening routeStarting point: Several desired targets, mutants, homologs, or anti-targets.
Primary output: Compound-by-target profiles and balanced candidates.
View this screening routeStarting point: A chemical space too large for uniform high-cost screening.
Primary output: A staged, traceable, and diverse experimental set.
View this screening routeAccess our diverse and high-quality compound libraries, ranging from bioactive molecules to fragment-based collections.
| Library category | Available collections |
|---|---|
| Bioactive Compound Libraries |
|
| Natural Product Libraries |
|
| Drug-Like Compound Libraries |
|
| Fragment Libraries |
|
Clarify the biological question, current evidence, target or ligand readiness, chemical-space options, experimental capacity, and the decision the final shortlist must support.
Select the structure-, ligand-, pharmacophore-, inverse-, fragment-, multi-target-, or data-driven route and agree on filters, controls, shortlist size, and success criteria.
Standardize protein states, pockets, cofactors, ligand protonation, stereochemistry, tautomers, conformers, and source identifiers using a documented protocol.
Apply fast compatibility and property filters first, then allocate docking, model inference, and higher-resolution calculations to progressively enriched subsets.
Compare orthogonal scores or models, inspect poses and interactions, cluster chemistry, and review property, liability, novelty, and procurement constraints.
Select a diverse primary test set and backups, document the rationale and limitations, and define the biochemical, biophysical, or cellular assays needed next.
When experimental results are available, interpret active and inactive outcomes, refine the screening hypothesis, and prioritize analogs or the next screening cycle.
Virtual screening prioritizes candidates; it does not prove binding or activity. We can help translate the computational result into an assayable compound set and an evidence-building plan.
Enzyme inhibition, receptor binding, kinase or protease activity, and target-specific assay planning.
Orthogonal binding or stability methods selected for the target, compound, and expected affinity range.
Pathway, reporter, viability, antiviral, phenotypic, target-engagement, or dose-response readouts where appropriate.
Related-target comparison, anti-target review, cytotoxicity controls, and confirmation of a cleaner activity profile.
Follow-up support can also include molecular dynamics, binding-energy analysis, in silico ADMET assessment, hit confirmation, analog expansion, and integrated interpretation of computational and experimental results.
Target class and disease context determine the relevant structures, chemical space, filters, assays, and acceptable risk profile. These factors are incorporated during project scoping.
We choose the route from input quality and the project decision rather than forcing every program through the same pipeline.
Target, compound, score, filter, and exclusion records stay traceable from source library to final shortlist.
Final nominations balance computational evidence with diversity, properties, alerts, novelty, and practical availability.
Outputs are connected to validation, counter-screening, analog expansion, and the next discovery decision.
Use structure-based screening when a qualified target structure and site are available; ligand-based or pharmacophore screening when reliable active molecules are the strongest evidence; inverse screening when the compound is known but the target is not; fragment screening for small, ligand-efficient starting points; multi-target screening for selectivity or polypharmacology; and AI-assisted high-throughput screening for very large chemical spaces. We can combine methods when the evidence supports it.
A useful first discussion covers the scientific question, target or ligand evidence, species and disease context, available structures or activity data, the compound source, experimental test capacity, timeline, and the decision the shortlist must support. Partial inputs are acceptable; feasibility work can fill defined gaps.
Yes, after checking confidence and local pocket quality. A predicted model may require refinement, ensemble preparation, or orthogonal ligand-based evidence. If the site is not suitable for docking, we explain the limitation and propose another route.
Yes. We can screen customer-supplied structures, selected commercial collections, or a defined public or proprietary chemical space. Structures are standardized and identifiers are preserved so results remain traceable.
Typical deliverables include a ranked and annotated compound list, screening-funnel statistics, method and parameter records, pose or feature evidence, diversity and property analysis, limitations, and a recommended experimental shortlist. Exact contents are agreed at scoping.
No. Virtual screening prioritizes hypotheses and reduces the number of compounds that need to be tested. Binding, activity, selectivity, and mechanism require experimental confirmation.
Yes. The computational shortlist can be paired with biochemical, biophysical, and cell-based testing, followed by hit confirmation, counter-screens, or iterative model updates.
Project inputs, target information, compound structures, and results can be handled under an appropriate confidentiality agreement and a project-specific data plan.
Talk to our technical team about your project!
I Want To Talk
