IPR Strategies For Synthetic-Biology Therapeutics And AI-Assisted Design.
IPR Strategies for Synthetic Biology Therapeutics and AI-Assisted Design
I. Introduction
Synthetic biology therapeutics combine biotechnology, genetic engineering, and chemical biology to design novel therapeutics. AI-assisted design accelerates drug discovery, protein engineering, and metabolic pathway optimization.
Key IP challenges include:
Patent Eligibility: Can AI-designed molecules, gene sequences, or synthetic organisms be patented?
Ownership Issues: Who owns AI-generated inventions?
Trade Secret Protection: Protecting proprietary algorithms, cell lines, and metabolic designs.
Cross-Border Enforcement: IP laws vary across jurisdictions for biotech and AI inventions.
Regulatory Overlap: FDA, EMA, and other regulators may affect patent strategies.
II. Key Intellectual Property Areas
1. Patents
Patentable subject matter includes:
Synthetic genes or proteins (if human-made and novel)
AI-designed chemical compounds or metabolic pathways
Methods of using engineered organisms
Challenges:
Natural phenomena exception (Myriad Genetics, 2013)
Abstract idea / AI algorithms (Alice Corp., 2014)
Inventorship: whether AI counts as an inventor (Thaler v. Perlmutter, 2023)
2. Trade Secrets
Protect proprietary AI models, datasets, or synthetic biology design protocols.
Critical in fast-moving biotech where patent publication may lead to competitive disclosure.
3. Copyright
Generally protects software implementing AI-assisted design.
Does not protect AI-generated molecules or synthetic pathways.
III. Detailed Case Laws
1. Association for Molecular Pathology v. Myriad Genetics, Inc. (2013, U.S. Supreme Court)
Facts:
Myriad patented isolated BRCA1 and BRCA2 genes linked to breast cancer.
Issue:
Are naturally occurring genes patentable?
Judgment:
Naturally occurring DNA cannot be patented.
cDNA (synthetic DNA) is patentable because it is human-made.
Relevance:
For synthetic biology therapeutics:
AI or human-designed DNA, RNA, or proteins can be patented.
Purely natural sequences cannot.
IP Strategy:
Focus patents on synthetic sequences, modified pathways, or engineered proteins rather than natural genes.
2. Diamond v. Chakrabarty (1980, U.S. Supreme Court)
Facts:
Chakrabarty engineered a genetically modified bacterium that could degrade oil spills.
Issue:
Are genetically modified organisms patentable?
Judgment:
Yes, human-made organisms are patentable.
“Anything under the sun that is made by man” is eligible.
Relevance:
Synthetic organisms and AI-designed microbes for therapeutics fall under this principle.
Patents can protect novel microbial strains and engineered cells.
IP Strategy:
Combine AI-assisted design with patentable synthetic organisms.
Claim both composition of matter and methods of use.
3. Thaler v. Perlmutter (2023, U.S. District Court)
Facts:
Stephen Thaler applied for copyright for works autonomously generated by AI (“Creativity Machine”).
Issue:
Can AI be an inventor or author under U.S. law?
Judgment:
Only humans can be recognized as inventors or authors.
AI cannot hold IP rights independently.
Relevance:
AI-assisted design can be patentable only if a human inventorship is present.
Human inventors must exercise creative control over AI outputs.
IP Strategy:
Clearly document human contributions in AI-assisted synthetic biology inventions.
List human inventors on patent filings.
4. Mayo Collaborative Services v. Prometheus Laboratories, Inc. (2012, U.S. Supreme Court)
Facts:
Prometheus patented a method linking metabolite levels to drug dosage.
Issue:
Are natural correlations patentable?
Judgment:
Laws of nature cannot be patented.
Must include innovative, inventive steps beyond natural principles.
Relevance:
AI-designed biomarkers or metabolic pathways must demonstrate novel, non-obvious modifications.
AI suggesting known natural correlations is not patentable.
IP Strategy:
Emphasize engineering or optimization steps in AI-assisted pathways.
Claim practical, technical applications, not natural phenomena.
5. Alice Corp. v. CLS Bank (2014, U.S. Supreme Court)
Facts:
Alice Corp. claimed software-based methods for financial transactions.
Issue:
Are computer-implemented abstract ideas patentable?
Judgment:
Abstract ideas implemented on a computer are not patentable.
Must show “significantly more” technical implementation.
Relevance:
AI algorithms for drug design or synthetic biology cannot be patented in isolation.
Must be tied to specific technical applications (e.g., DNA synthesis device, robotic lab system).
IP Strategy:
Combine AI methods with hardware or lab automation in patent claims.
Avoid claiming pure software/mathematical algorithms.
6. Waymo LLC v. Uber Technologies, Inc. (2017, U.S. District Court)
Facts:
Waymo claimed Uber stole AI trade secrets for self-driving technology.
Judgment:
Trade secret protection applies to confidential AI algorithms and datasets.
Misappropriation leads to significant remedies.
Relevance:
AI-assisted synthetic biology platforms rely on proprietary datasets, models, and design protocols.
Trade secrets complement patent protection, especially for fast-moving technology.
IP Strategy:
Keep AI models, training data, and metabolic design protocols confidential.
Implement strict access controls and NDAs.
7. Biogen v. Mylan (2020, U.S. District Court)
Facts:
Biogen patented biosimilar formulations; Mylan challenged patent validity.
Judgment:
Patents on specific engineered protein sequences are enforceable if inventive.
Generic challenges focus on obviousness or prior art.
Relevance:
AI-designed therapeutics must be novel and non-obvious.
Detailed patent drafting is required to capture unique molecular modifications.
IP Strategy:
Patent composition of matter and methods of use for AI-engineered proteins.
Include supporting data from AI simulations or lab validation.
IV. Strategic IP Approaches for AI and Synthetic Biology Therapeutics
| IP Type | Strategy |
|---|---|
| Patents | Claim synthetic genes, proteins, or pathways; include method-of-use claims; integrate AI design with lab automation. |
| Trade Secrets | Protect AI models, training datasets, cell line designs, and synthetic protocols; implement NDAs and access controls. |
| Copyright | Protect AI software and source code; content generated by AI without human inventorship may not be copyrightable. |
| Documentation | Maintain clear records of human involvement in AI-assisted design. |
| Global Enforcement | Consider cross-border patent filings (PCT), regional trade secret protection, and regulatory compliance. |
V. Key Takeaways from Cases
Human inventorship is mandatory (Thaler v. Perlmutter).
Synthetic biology inventions are patentable if engineered or designed by humans (Chakrabarty, Myriad).
AI alone cannot claim IP rights, but AI can augment human inventors.
Trade secrets are critical for proprietary AI models (Waymo v. Uber).
Abstract algorithms or natural phenomena are not patentable (Mayo, Alice).
Technical implementation and integration with lab systems strengthen patent eligibility.
VI. Conclusion
IP strategy for synthetic biology therapeutics and AI-assisted design requires:
Patents: Protect engineered sequences, molecules, pathways, and methods, not just algorithms.
Trade Secrets: Safeguard AI models, datasets, and experimental protocols.
Human Inventorship Documentation: Essential for AI-assisted inventions.
Integration with Technical Systems: Improves patentability and enforcement.
Cross-Border Considerations: File PCT applications and ensure compliance with local biotech/IP regulations.
RELATED Blog
comments