For technology companies that patent foundational technology while continuing confidential research and development, the difficult question is not whether a patent disclosure is public. It is how far that disclosure reaches. The District of Delaware’s September 3, 2026 decision in Praxis Precision Medicines, Inc. v. Goldstein, No. 1:25-cv-01256-JCG (D. Del. Sept. 3, 2026), illustrates why that boundary may resist early resolution when a company can identify a confidential layer that allegedly goes beyond its published patent application and scientific literature.
Praxis alleged that former scientific advisor David Goldstein disclosed or used confidential KCNT1 inhibitor information in connection with Actio Biosciences’ competing drug-development program. The defendants argued that the alleged secrets were not secret because Praxis’s own 2020 international patent application and 2021 scientific article already disclosed the relevant family of compounds and the kinds of chemical modifications at issue. The court denied the motion to dismiss, concluding that the present record did not establish that Praxis’s 14 alleged confidential compounds were readily ascertainable from the public disclosures and that the complaint plausibly alleged trade-secret misappropriation and breach of contract.
The limits of that ruling matter. Judge Jennifer Choe-Groves, a judge of the U.S. Court of International Trade sitting by designation in the District of Delaware, did not find that Praxis ultimately owns valid trade secrets, that Goldstein disclosed them, or that Actio copied them. The decision allows the claims to proceed to discovery and fact-finding. It is a nonbinding district-court ruling whose substantive importance derives principally from its application of binding Third Circuit decisions, particularly Mallet & Co. v. Lacayo, 16 F.4th 364 (3d Cir. 2021), and Oakwood Laboratories LLC v. Thanoo, 999 F.3d 892 (3d Cir. 2021).
The Dispute Was About What Remained Secret After Praxis Published
Praxis is a clinical-stage biopharmaceutical company developing therapies for central nervous system disorders. According to the complaint, its KCNT1 program generated non-public candidate compounds, information about their identities, formulas, structures, properties, and inhibitory activity, as well as non-public assay results, comparative analyses, and related know-how. Praxis alleged that Goldstein, a Lead Scientific Founder and Chairman of its Scientific Advisory Board, had broad access to that information and was bound by confidentiality obligations under a Founder and Scientific Advisory Board Agreement and a later Master Advisory Agreement.
Praxis further alleged that Goldstein helped establish Actio Biosciences while he remained affiliated with Praxis and that Actio later pursued its own KCNT1 inhibitor program. When Actio patent applications became public in 2025, Praxis alleged that compounds disclosed in those applications bore striking structural similarities to 14 non-public Praxis compounds. Those allegations, including side-by-side molecular comparisons reproduced in the court’s opinion, formed the factual basis for Praxis’s inference that its confidential compounds had been used in Actio’s development work. The defendants denied that inference and maintained that Actio independently developed its compounds from public information.
The defense gained force from Praxis’s own publication history. Praxis’s international patent application, KCNT1 Inhibitors and Methods of Use, was published in November 2020 and disclosed a family of KCNT1 inhibitor compounds and numerous structural variations. In March 2021, Praxis researchers also published an article by Andrew Griffin and colleagues in ACS Medicinal Chemistry Letters describing an oxadiazole series of KNa1.1 inhibitors, including Compound 31, along with structure-activity relationships and experimental data across multiple compounds.
The Core Question Was How Far the Public Disclosures Reached
The defendants did not argue that Praxis had literally published each of the 14 compounds it later identified as confidential. Their argument was broader: the public patent application and scientific article had disclosed enough of the relevant chemical space that the claimed confidential compounds were readily ascertainable through proper means and therefore could not qualify as trade secrets under the Defend Trade Secrets Act.
That argument exposed an important tension in Praxis’s theory. To support an inference of misappropriation, Praxis alleged that certain Actio compounds differed only modestly from its confidential compounds and characterized some of the relevant modifications as commonplace for medicinal chemists. The defendants tried to turn that allegation against Praxis. If those modifications were routine, they argued, the same logic suggested that Praxis’s undisclosed compounds could be derived from the public Compound 31 and related disclosures without access to any secret information.
Praxis responded that the comparisons were not symmetrical. It argued that materially greater changes were necessary to move from public Compound 31 to its confidential compounds, while some Actio compounds differed from the confidential Praxis compounds by relatively limited structural changes. Praxis also emphasized that the defendants’ detailed public-domain analysis focused on two of the 14 alleged confidential compounds rather than the full set. That distinction became central to the court’s refusal to resolve the issue at this stage.
Mallet Recognizes That Trade Secrets Can Extend Beyond Patent Disclosures
The court relied on Mallet & Co. v. Lacayo for the governing boundary. Mallet involved alleged trade secrets in commercial baking release agents and arose from a preliminary injunction that the Third Circuit ultimately vacated because the claimed secrets and the injunction were not defined with sufficient precision. In discussing what could qualify as secret, however, the Third Circuit made clear that information actually disclosed in patents or otherwise made public cannot remain a trade secret, while proprietary information that goes beyond the patent disclosure or refines the disclosed process may remain protectable.
That principle is narrower than saying patents and trade secrets always coexist. The trade-secret owner must still identify information that was not publicly disclosed or rendered readily ascertainable through proper means and must satisfy the statutory requirements of secrecy, economic value derived from secrecy, and reasonable measures to maintain confidentiality. In Praxis, the court concluded only that, at this preliminary stage, the Griffin article and the KCNT1 patent application did not establish that the alleged confidential compounds were readily ascertainable.
The breadth of the defendants’ showing also mattered. The complaint identified 14 confidential Praxis compounds, while the defendants’ specific comparison concentrated on two. Praxis, in turn, offered its own examples attempting to show why those compounds were not readily ascertainable from Compound 31. Against that record, the court treated the scientific relationship between the public and confidential compounds as a factual question requiring further development rather than a point that could conclusively eliminate the claims at the outset.
Praxis Also Identified the Alleged Trade Secrets With Enough Specificity to Plead a Claim
The court separately applied Oakwood Laboratories LLC v. Thanoo, the Third Circuit’s controlling pleading authority for DTSA claims. Oakwood rejects both extremes that frequently distort trade-secret pleading. A plaintiff cannot rely on broad labels that leave the defendant guessing what information is claimed as secret, but it also need not disclose the secret in such detail that filing suit destroys the very confidentiality the statute protects.
Praxis alleged more than an undifferentiated category such as drug-development know-how. It identified non-public candidate KCNT1 compounds and information concerning their identities, formulas, structures, properties, and inhibitory activities, together with non-public assay and comparative-analysis results and additional know-how. It also alleged concrete secrecy measures, including confidentiality agreements, data classification, employee training, a code of conduct, and IT security controls. The court held that those allegations were sufficiently particular, at the pleading stage, to distinguish the claimed information from general knowledge and put the defendants on notice of the claims.
That conclusion should not be confused with final trade-secret identification for summary judgment or trial. Oakwood is pleading-stage authority, and Praxis does not establish the degree of specificity that will be required after discovery, at summary judgment, or at trial. For technology companies, the practical lesson is to maintain an internal disclosure map that can distinguish public patent claims, specifications, papers, presentations, product releases, and other disclosures from the later data, parameters, implementations, failed experiments, and optimization work the company intends to keep confidential.
A Separate Procedural Question: Was Trade Secret Status Properly Treated as Jurisdictional?
The opinion contains a procedural issue that is easy to miss but legally significant. The defendants did not limit their challenge to Rule 12(b)(6). They also mounted what the court characterized as a factual Rule 12(b)(1) attack, arguing that Praxis lacked a protectable DTSA trade secret and therefore lacked a federal claim capable of supporting subject-matter jurisdiction. Applying Third Circuit decisions governing factual jurisdictional challenges, the court considered evidence outside the pleadings and placed the burden on Praxis to establish jurisdiction.
That treatment deserves caution. The DTSA separately provides in 18 U.S.C. § 1836(c) that federal district courts have original jurisdiction over civil actions brought under the statute, while § 1836(b)(1) creates the private cause of action. Oakwood describes the existence of a trade secret as an element of a DTSA claim. The Supreme Court’s decision in Arbaugh v. Y & H Corp., 546 U.S. 500 (2006), instructs courts generally to treat statutory limitations as nonjurisdictional unless Congress clearly ranks them as jurisdictional, in part to avoid converting merits disputes into threshold questions of judicial power.
The Praxis opinion did not discuss Arbaugh or separately analyze whether the existence of a protectable trade secret is itself a jurisdictional prerequisite rather than a merits element. Because the court denied the Rule 12(b)(1) motion and then independently concluded under Rule 12(b)(6) that Praxis had plausibly alleged DTSA and Delaware Uniform Trade Secrets Act claims, the characterization did not terminate the case. Still, practitioners should treat this part of the decision as a procedural wrinkle in a nonbinding district-court opinion, not as a broad holding that courts may ordinarily resolve the existence of a DTSA trade secret as a factual jurisdictional question before discovery.
Access and Technical Similarity Supported Plausibility, Not a Finding of Copying
Praxis’s misappropriation theory was circumstantial. It alleged that Goldstein had extensive access to the confidential KCNT1 information, was subject to contractual confidentiality duties, remained affiliated with Praxis while helping form Actio, and that Actio’s later patent applications disclosed compounds allegedly resembling Praxis’s non-public compounds. The side-by-side molecular comparisons gave those allegations a technical basis beyond the mere fact that Goldstein later became a competitor.
The defendants offered lawful alternatives. They argued that Goldstein was not identified as an inventor on the relevant Actio applications, that the similarities were not probative of copying, and that Actio could have developed its compounds independently from Praxis’s public disclosures and routine medicinal-chemistry methods. The court did not decide between those explanations. Under the Rule 12(b)(6) standard, it held that the alleged access, confidentiality obligations, timing, and technical comparisons plausibly supported disclosure and use under the DTSA and the Delaware Uniform Trade Secrets Act.
The distinction is important. The ruling does not establish that molecular resemblance proves misappropriation, does not require Actio to prove independent development, and does not resolve whether the alleged confidential compounds were actually used in Actio’s research. On the allegations presented, the defendants’ independent-development theory did not justify ending the case before discovery.
The Contract Claims Provide a Parallel Theory, But the Decision Is Narrow
Praxis also asserted two Delaware-law contract claims against Goldstein under the Master Advisory Agreement. Section 6 allegedly required Goldstein to protect and limit his use of Praxis confidential information and materials, while Section 5 allegedly addressed ownership and assignment of work product. Applying Delaware contract law, the court held that Praxis plausibly alleged the existence of the MAA, breach of Goldstein’s contractual obligations, and resulting damages.
For companies that depend on founders, scientific advisors, consultants, contractors, and collaboration partners, the overlap is commercially important. Trade-secret law imposes statutory requirements that exist independently of contract, including secrecy, economic value derived from secrecy, and reasonable protective measures. Contract terms can separately define permitted uses, confidentiality duties, treatment of company materials, and ownership of work product, which can reduce uncertainty when a relationship ends or a participant later joins a competing effort.
The decision should not be read more broadly than that. It did not decide the ultimate scope or enforceability of the MAA, whether any particular Actio invention qualifies as Praxis work product, or how contractual confidentiality obligations would apply to information proven to be public. It held only that Praxis’s allegations under Sections 5 and 6 were sufficient to proceed alongside the statutory trade-secret claims.
Practical Implications for R&D-Intensive Technology Companies
The most useful lesson from Praxis is not that later R&D automatically survives a patent disclosure. It is that companies pursuing patents and trade secrets simultaneously should be able to identify the boundary between them with technical precision. Before a patent application, paper, conference presentation, open-source release, benchmark, regulatory submission, or product launch becomes public, the company should identify what information is being disclosed and what related information is intended to remain confidential.
That exercise should continue after publication. In software, the public disclosure may describe an architecture while training data, tuning parameters, deployment methods, failure analysis, or performance results remain confidential. In hardware and manufacturing, public claims may coexist with confidential tolerances, process windows, test data, tooling methods, yield improvements, and supplier-specific know-how. In life sciences, published compound families or target mechanisms may coexist with non-public candidate selection, assay results, optimization paths, negative data, and later-developed structures. The legal question is not whether the confidential information is related to the public technology, but whether it remains genuinely distinct, valuable because of secrecy, and not readily ascertainable through proper means.
The litigation consequence is equally concrete. A company that has already made substantial public disclosures will be vulnerable if it later describes its trade secrets only as additional know-how. The stronger position is to be able to show, before litigation begins, what the public disclosure taught, what remained internal, when the internal information was developed, who had access to it, and what confidentiality restrictions applied. That record helps address both sides of a future dispute: whether the information was truly secret and whether similarities in a competitor’s work are more consistent with misuse than with independent development from the public domain.
Key Takeaways
- Patent publication does not automatically eliminate every related trade secret. Information actually disclosed is public, but later or additional information may remain protectable if it goes beyond the disclosure and satisfies the statutory requirements for secrecy.
- The public/private boundary must be technically identifiable. A trade-secret owner should be able to distinguish the claimed confidential layer from patents, papers, product releases, presentations, and other public information.
- Readily ascertainable may be fact-intensive. Where readily ascertainable turns on competing technical comparisons and the plaintiff has identified specific non-public information, Praxis shows why the issue may resist resolution at the pleading stage.
- Specificity matters from the outset. Under Oakwood, the plaintiff must identify the boundaries of the alleged secrets without pleading them so comprehensively that the complaint destroys their secrecy.
- Access plus technical similarity can support a plausible inference of misappropriation, but it is not proof of copying. Independent development remains a merits defense, and the plaintiff retains the ultimate burden of proving misappropriation.
- Contractual protections can complement trade-secret law. Clear confidentiality, use, material-return, work-product, and assignment provisions can create an additional contractual framework for sensitive R&D relationships, although Praxis did not finally determine the scope of those obligations.
- The precedential significance is limited. Praxis is a nonbinding District of Delaware motion-to-dismiss ruling, while the principal binding guidance comes from Third Circuit decisions such as Mallet and Oakwood.
Related Analysis
- International Medical Devices v. Cornell provides the contrasting case in which prior patent disclosures did defeat claimed trade-secret protection because the asserted concepts were already public.
- Pegasystems v. Appian examines the consequences of failing to identify trade secrets with sufficient specificity at later stages of litigation and provides a useful counterpoint to Oakwood‘s pleading-stage standard.
- Insulet v. EOFlow addresses circumstantial access-plus-similarity evidence in trade-secret litigation and shows how the same kinds of facts can become important not only to pleading misappropriation but also to statute-of-limitations analysis.