Design Around Patents

​

The Problem You Are Trying to Solve

“I have a set of lead compounds that infringe on existing patents. I need to design novel chemical matter that preserves favorable performance while avoiding infringement.” At the late lead optimization stage, patent risk becomes a gating constraint. Even highly effective compounds may be unusable if they:

• Fall within protected scaffolds or claim language
• Are obvious analogs under doctrine-of-equivalents reasoning
• Depend on protected substitutions, linkers, or core motifs

This workflow helps you systematically escape patent space while maintaining on-target performance, using physics-based validation and AI-guided redesign.

​

Solution

This workflow centers on controlled chemical novelty:

• Preserve functional performance (binding mode, interactions, properties)
• Alter chemical identity (scaffold, connectivity, electronics, shape) enough to establish novelty

The primary redesign chain is: Patent-Aware Deconstruction → Motif Preservation → Novel Chemistry Generation → Performance Re-validation → Developability & Synthesis Checks. Binding, quantum chemistry, and AI engines work together to ensure replacements are both non-infringing and credible leads.

​

What Data Do I Need to Provide?

Required

• Lead compounds as SMILES (the infringing set)
• Target protein structure (experimental or modeled)

Recommended

• Knowledge of why the compound performs well (binding mode, key residues)
• Known patent boundaries (claimed scaffolds, functional groups, substitution rules)

Optional

• Internal SAR or experimental benchmarks
• Property constraints (ADMET, solubility, toxicity ceilings)

​

Workflow

1. Deconstruct the Infringing Leads

Before changing chemistry, identify what cannot be lost. Users typically:

• Analyze docking poses or experimental complexes
• Identify key interaction motifs (H-bond donors/acceptors, hydrophobic anchors, π-stacking regions)
• Separate performance-critical features from chemically incidental features
• By understanding structure activity relationships and critical motifs that drive favorable binding or other properties, you can ensure preservation of the key scaffolds on the molecules during expansion efforts.

Primary engines used:

• Docking / Ensemble Docking to visualize conserved binding modes
• Protein-Ligand MD to confirm which interactions are stable vs. incidental
• Pharmacophore Analysis to abstract interactions into claim-agnostic features
• QSAR Modeling to get a better picture of the chemical space and which substructures are driving favorable properties of your leads

This step answers:

• What makes this molecule work?
• Which elements must be preserved in function, not structure?

This step will give you a functional interaction map and motif definition, independent of patented chemistry. We would recommend taking this key scaffold and running an independent search on the patent databases like sureChemBL.

2. Map the Patent Risk Surface

With functional motifs defined, shift focus to where you cannot go. Users typically:

• Identify scaffold cores, linkers, or substituent patterns that overlap claims
• Flag chemical regions that are too close to existing exemplified compounds
• Decide whether novelty should come from:
  • Scaffold hopping
  • Conformational reshaping
  • Complete reconstruction of the molecule

This step informs how aggressive the redesign must be, providing clear “no-go” regions and acceptable novelty strategies.
You can begin this process by throwing your compound into SureChemBL, searching for infringing patents, and diving into the details to know the breadth of patents for your critical scaffolds.

3. Generate Novel Chemical Matter

Use Generative Chemistry engines to design new molecules that:

• Preserve pharmacophore features
• Replace patented scaffolds or connectivity
• Explore chemically distinct regions of space

Primary engines:

• De Novo Library Generation for scaffold-level novelty
• Molecular Optimization for controlled, constraint-aware redesign, geared towards property optimizations
• Custom Model Training (optional) to bias generation toward internal success criteria

Generation is typically constrained by:

• Required interaction features (from Step 1)
• Property and synthesizability bounds

This will provide novel, patent-diverse compound libraries aligned to the same biological objective. You will need to eventually re-screen your generated set of molecules into SureChemBL for your high priority lead molecules.

4. Re-Validate Performance Against the Target

Novelty alone is not enough; new compounds must still work. Primary engines:

• Docking → Flexible / Ensemble Docking to confirm binding feasibility
• Protein–Ligand MD to test pose stability and interaction persistence
• Free Energy Perturbation (RBFE / ABFE) to quantify performance relative to the original lead

This step answers:

• Do these new molecules bind the same way (functionally)?
• Have we preserved or improved potency?

This will result in a ranked set of non-infringing candidates with validated on-target engagement.

5. Evaluate Novelty-Driven Property Risk

Structural novelty can introduce new liabilities. Users typically assess:

• ADMET-AI for toxicity, metabolism, clearance, and safety flags
• Compound Solubility to avoid formulation regressions

Other Quantum Properties to ensure that the structure of the molecule has the right geometries, electronic properties, and metabolic profiles.
This step ensures novelty did not come at the cost of developability, giving a refined shortlist of viable, novel leads.

6. Confirm Buildability and Freedom-to-Operate (Optional)

Before committing, ensure redesigned compounds are practical. Optional next steps include:

• Retrosynthesis to verify synthetic accessibility
• SureChemBL patent database to search the novel structures for infringing patents based on tanimoto similarities and other criteria
• You can utilize all the other engines we offer to re-screen your new molecules to maintain favorable property profiles.

This will produce synthesis-ready, patent-aware lead candidates.

​

Results

• Chemically distinct leads that preserve functional performance
• Reduced patent infringement risk through scaffold and topology changes
• Quantitative confidence that performance has been retained or improved
• A clear audit trail from patented lead → novel candidate

Now what?

• After receiving your lead set, make sure to review the new compounds you’d like to take into synthesis into the SureChemBL Engine which will allow you to confirm patentability of your molecules
• You can re-screen all of your molecules using the rest of the Revilico Engines for properties you are interested in.

​

Integration with Other Engines (Optional)

This workflow integrates naturally with:

• Multi-parameter optimization pipelines (potency, ADMET, solubility)
• Toxicity and off-target screening workflows
• Experimental planning via prioritized synthesis routes

​

Why Revilico?

Revilico uniquely enables patent-aware lead redesign by combining:

• Generative chemistry with hard biological and chemical constraints
• Physics-based validation (MD, FEP) to protect performance
• Quantum and property engines to manage novelty risk
• AI-assisted interpretation to accelerate decision-making

This allows teams to move beyond “cosmetic analogs” and confidently develop truly novel chemical matter that stands up scientifically, commercially, and legally.