Insilico Medicine: Lessons in IP strategy from a front-runner in AI-drug discovery

The intersection of AI and pharmaceutical development presents unprecedented opportunities but also raises complex legal questions. Recent developments and successes in AI-drug discovery highlight some of the key IP issues in AI-drug development. Companies are being forced to tackle these issues head-on as the IP law advances almost as quickly as the science. In a previous post, this Kat took a look at the challenges for Google DeepMind's spin-out Isomorphic in protecting and commercialising its protein fold algorithm AlphaFold (IPKat). In this post, we will look at Insilico Medicine, another high-profile player in AI-drug discovery that is much further along in the drug discovery process. 

AI-assisted drug discovery

The landscape of drug discovery is undergoing a fundamental transformation. AI models are dramatically accelerating the traditionally time-consuming process of drug development (IPKat). These technologies enable researchers to predict molecular interactions and identify promising drug candidates with unprecedented accuracy and speed. However, whilst the potential benefits are substantial, the path to protecting AI innovations in drug discovery presents unique challenges. The rapid pace of development in the field means that companies must carefully navigate between maintaining competitive advantages, ensuring strong IP protection and participating in the broader scientific discourse (IPKat).

Kat-assisted AI drug discovery

Patents versus marketing in AI drug discovery

A crucial consideration in AI-assisted drug discovery is the relationship between patent protection and disclosure. Patents for new therapeutic compounds do not necessarily require disclosure of the discovery process and instead general focus on the properties and synthesis protocols for the compound. The key legal test is whether a skilled person could perform the invention. It is thus generally not necessary to describe how the invention was first derived. Nonetheless, the inventive story behind a novel compound may still play a crucial role during patent prosecution and/or subsequent litigation. When proposed compounds share structural similarities with the compounds of the prior art, the inventors will need to demonstrate the innovative aspects and superior properties of the invention. In this context, the use of AI in the discovery process might actually provide compelling evidence of non-obviousness, particularly when the AI system identifies unexpected benefits or novel applications. 

Insilico Medicine: AI-driven drug discovery

Insilico Medicine is one of the drug-discovery companies furthest ahead in terms of AI-designed clinical candidates. Since introducing generative AI for molecular design in 2016, the company has established itself as a leader in AI-driven drug discovery. Insilico's Pharma.AI platform encompasses a suite of tools spanning biology, chemistry, and clinical development. Insilico has so-far received IND approval for 10 molecules, and currently have drug candidates in phase I and II clinical trials. 

It appears that Insilico has taken a dual-approach to IP strategy. The company has granted patents relating to both its AI platform technologies (see e.g. US11403521 B2) and the composition matter (i.e. chemical identity) of its clinical candidates. 

Insilico's TNIK Nature Biotechnology paper

Last year, Insilico published a paper in Nature Biotechnology on the development of its most advanced drug candidate, the TNIK inhibitor INS018_055 (also known as ISM001-055). INS018_055 is a small-molecule inhibitor targeting TRAF2- and NCK-interacting kinase (TNIK) for treating fibrosis. According to the paper, the drug was developed using Insilico's AI-driven target discovery platform. First, using Insilico's PandaOmics platform, the authors identified TNIK as a promising anti-fibrotic target. Insilico's Chemistry42 AI platform was then used to design INS018_055 as a selective TNIK inhibitor. According to the paper, the development process from target discovery to preclinical candidate took only 18 months.

The paper goes on to describe preclinical studies in which INS018_055 demonstrated significant anti-fibrotic activity across multiple organs including lung, kidney, and skin. In animal models of lung fibrosis, kidney fibrosis, and skin fibrosis, INS018_055 treatment was shown to significantly reduce fibrotic tissue formation and improve organ function.

The paper also describes the results of two Phase I clinical trials (NCT05154240 and CTR20221542) evaluating INS018_055's safety and pharmacokinetics in healthy volunteers. According to the paper, the trials demonstrated that INS018_055 was safe and well-tolerated with good oral bioavailability and dose-proportional pharmacokinetics. Since publication of the paper, Insilico has announced that INS018_055 has also shown safety and efficacy in a subsequent phase IIa study. 

Insilico's TNIK inhibitor patent

A search of the public databases reveals that Insilico has a granted US patent for TNIK inhibitors. US11530197 B2 describes compounds designed to treat fibrotic diseases. The PCT publication is, in fact, referenced in the Nature Biotech paper. The compounds disclosed in the patent are characterised through various biological assays including human liver microsome stability tests, TNIK and MAP4K4 enzyme inhibition assays, cell-based fibrosis studies, and animal models of lung and skin fibrosis. Compound 112 (claim 5) of the patent appears to correspond to INS018_055. According to the patent, Compound 112 showed promising results in terms of human liver microsome stability and TNIK inhibition. The patent also provides a detailed synthesis procedure for Compound 112 and suggests that the overall profile of this compound suggests potential therapeutic utility in treating fibrotic conditions like idiopathic pulmonary fibrosis.

The patent does not explicitly describe how the compounds described therein were initially identified. Whilst the Examples provide extensive details about the synthesis methods and biological testing of the compounds, it does not reveal the drug discovery strategy or screening approach that led to these particular molecular structures. However, this is not unusual for a small molecule composition of matter patent. It is not necessary for patentability for the discovery processes leading to an invention to be disclosed. Nonetheless, the lack of this disclosure in the patent is in sharp contrast with the Nature Biotechnology paper

To disclose or not disclose? 

Given that Insilico has chosen to disclose the discovery process leading to INS018_055 in an academic publication, why then did they not include these details in the patent? Given that the US patent has already been granted, it is clear that the US patent office at least was not concerned by the lack of details about the discovery process, in line with the established case law. However, inclusion of the discovery processes would arguably have benefited the patent. For example, in jurisdictions such as Europe in which the inventive story behind the claimed invention plays a big part in validity, the evidence from the AI-discovery process may have strengthened the filing and may even have permitted a broader scope of protection by providing the evidence and inventive argument behind a broader invention. 

On the other hand, there may also be dangers in divulging your drug discovery process. Given the growing number of AI-drug discovery companies and patents, you may for example, reveal inadvertent patent infringement by describing your drug development process (one particularly thinks of the AlphaFold patents). There is also a possibility that inventorship of the patent may be challenged, depending on the relative contributions of the human and AI to the invention (IPKat). Companies must now weigh up the risks and benefits of disclosure. It is likely that traditional pharma companies will choose not to disclose the secret sauce the lead to the discovery of their pharmaceutical inventions. However, for Companies such as Insilico Medicine who are entirely focused on AI-drug discovery, and are clearly seeking to foster collaborations with traditional pharma, the benefits of publishing their AI success stories may outweigh any concerns about disclosure. 

Final thoughts

The evolving landscape of AI-assisted drug discovery presents both opportunities and challenges for IP strategy. The balance between open science, increasing a company's profile and ensuring strong IP protection for the company's assets remains delicate. As the field continues to mature, it will be interesting to see whether the traditional approaches to composition of matter patent drafting evolves to include more information about technologies involved in discovery, testing and validation.

Further reading

Insilico Medicine: Lessons in IP strategy from a front-runner in AI-drug discovery Insilico Medicine: Lessons in IP strategy from a front-runner in AI-drug discovery Reviewed by Rose Hughes on Tuesday, February 04, 2025 Rating: 5

No comments:

All comments must be moderated by a member of the IPKat team before they appear on the blog. Comments will not be allowed if the contravene the IPKat policy that readers' comments should not be obscene or defamatory; they should not consist of ad hominem attacks on members of the blog team or other comment-posters and they should make a constructive contribution to the discussion of the post on which they purport to comment.

It is also the IPKat policy that comments should not be made completely anonymously, and users should use a consistent name or pseudonym (which should not itself be defamatory or obscene, or that of another real person), either in the "identity" field, or at the beginning of the comment. Current practice is to, however, allow a limited number of comments that contravene this policy, provided that the comment has a high degree of relevance and the comment chain does not become too difficult to follow.

Learn more here: http://ipkitten.blogspot.com/p/want-to-complain.html

Powered by Blogger.