The patent covers the multi-dimensional biological representation layer organized around MindWalk’s HYFT® pattern anchors — the model-agnostic context layer designed to support ReefIQ™, LensAI™, and customer-selected AI workflows as they retrieve, compare, and reason over connected, traceable biology in drug discovery.
AUSTIN, Texas–(BUSINESS WIRE)–$HYFT #ai—MindWalk Holdings Corp. (“MindWalk”) (NASDAQ: HYFT), a Bio-Native AI company, today announced the filing of European patent application No. EP26187897.9, directed to high-dimensional data structures for biological subsequences and property inference. The application is intended to protect the enriched biological representation architecture that underpins MindWalk’s HYFT® Technology, its ReefIQ™ biological context layer, and its LensAI™ reasoning workflows.
The filing comes as agentic AI in life sciences converges on a clear architectural lesson: frontier models alone are not enough. Recent public work on scientific agents, including NVIDIA’s BioNeMo Agent Toolkit [1] and AstraZeneca’s ChatInvent system [2], points to the need for domain-specific context, structured tool interfaces, reliable input/output schemas, provenance, memory, and validation when AI agents are deployed in real scientific workflows. As frontier models trend toward commodity, MindWalk believes the durable advantage migrates to the data layer, the structured, domain-specific representation that lets any model or agent retrieve, compare, and reason over biology with traceable context.
“As AI models become more broadly available, we believe lasting advantage in life sciences will come from the biological context those models can use. The differentiator is not a generic AI interface, but the structured biological representation that lets models and agentic workflows retrieve connected evidence, preserve provenance, and reuse knowledge across programs. This filing is intended to protect aspects of the architecture at the core of HYFT® Technology, where MindWalk believes lasting value in biological AI can compound.”
— Jennifer Bath, Ph.D., Chief Executive Officer and President, MindWalk Holdings Corp.
That data layer also anchors a large and growing market. Spending on AI in drug discovery is projected to grow from approximately US$5 billion in 2026 to more than US$8 billion by 2030 [3], and it sits atop the more than US$250 billion the pharmaceutical industry invests in research and development each year [4].
The new application builds on MindWalk’s foundational HYFT patent (EP3881326A1), which established how the Company identifies the characteristic biological patterns that recur across life and uses them as a searchable language to compare sequences without alignment. EP26187897.9 protects a distinct and additive layer: organizing the biological meaning around those patterns into a form that can be reused across MindWalk’s infrastructure, customer programs, and AI workflows. It is the computational layer built on the original foundation, not a re-filing of it.
MindWalk’s earlier HYFT® patent estate established the foundation of the Company’s biological intelligence platform: the ability to identify recurring biological patterns and use them as a searchable language across life sciences. The newly filed patent is intended to address a distinct, additive layer above that foundation — by seeking protection for how biologically meaningful subsequence-level pattern anchors can be enriched with connected biological context and made usable across MindWalk’s platform, customer programs, and AI workflows.
The added value of the new representation layer is that each HYFT® pattern can become more than a search result, annotation, or output from an AI model. The filing is intended to help keep the biological meaning around each pattern connected — including sequence-level, structural, physicochemical, functional, experimental, literature-derived, and program-history context — so that ReefIQ™, LensAI™, and customer-selected AI workflows can retrieve and reuse that knowledge over time.
This makes the layer complementary to protein LLMs and other AI models. Those models are powerful, but MindWalk believes lasting platform value comes from the biological context that models can use, not from any single model itself. By keeping biological meaning organized in a model-agnostic layer, the filing is intended to support a data asset that can become more valuable as more programs, evidence, and customer data are connected to the platform.
“Biology does not live in one file type. Sequence, structure, physicochemical behavior, function, evidence, and literature all need to stay connected for AI systems to be useful in discovery. Through this filing, we are seeking to protect the layer that keeps those biological connections organized, accessible, and usable as discovery work moves between AI models, software tools, and human scientific teams. That is the layer we believe can make AI workflows more grounded, more reusable, and more valuable over time.”
— Dirk Van Hyfte, M.D., Ph.D., Chief Technology Officer, MindWalk
The approach reflects MindWalk’s functional and evolutionary-constraint thesis. Biology is not random text: recurring patterns often persist because they are tied to function, structure, binding, immune recognition, or evolutionary fitness. By representing those patterns together with their surrounding biological context, MindWalk aims to give AI systems a more organized and traceable basis for reasoning over biology, not just more data, but better-connected biological meaning.
MindWalk has begun to demonstrate the approach in active programs. In dengue, the Company has reported preclinical binding-level data in which a HYFT®-identified target informed immunogen design and generated antibodies that bound across antigens from all four dengue serotypes in two independent campaigns. In influenza, MindWalk has disclosed a HYFT®-defined functional constraint observed across broad influenza A and B datasets, including human, avian, swine-associated, Victoria, and Yamagata backgrounds. These programs remain preclinical, and additional studies are required to assess neutralization, safety, durability, clinical translation, regulatory path, and commercial potential.
ReefIQ™ and LensAI™ are the commercial expression of this representation strategy. LensAI™ is in contracted, recurring arrangements with life-sciences customers today, and the filing is intended to protect a layer those deployments can build on as data, evidence, and program experience accumulate across the platform. As life-sciences organizations move toward more governed, AI-assisted discovery, MindWalk believes traceable biological context, reusable program memory, and model-independent access to structured biological knowledge will become increasingly important.
For investors, the significance is that the filing supports MindWalk’s strategy of building value beyond any single AI model. In that stack, HYFT® identifies biologically meaningful pattern anchors; ReefIQ™ organizes customer and program data around those anchors as governed biological context; and LensAI™ operates over that context to support target discovery, candidate diligence, hypothesis generation, and portfolio decision support. The durable asset is the biology-aware representation layer itself, a model-agnostic infrastructure asset that can become more valuable as more programs, evidence, and customer data are connected to it.
About MindWalk Holdings Corp.
MindWalk Holdings Corp. (NASDAQ: HYFT) is a Bio-Native AI company building the BioIntelligence infrastructure that AI-driven life sciences and agentic workflows require, integrating models, data, and advanced wet lab capabilities into one connected discovery ecosystem. At its core is HYFT® Technology, a proprietary, function-aware representation of biology whose HYFT fingerprints span sequence and structural biology, refined over 20 years of curation, form a continuously evolving biological representation of 660 million patterns and 25 billion relationships. This enriched biological representation is the architecture behind ReefIQ™, the biological context layer for life sciences, enriching data at ingestion and growing more valuable with each program run on it, and LensAI™, the reasoning and application layer for target discovery, candidate diligence, portfolio decision support, and the agentic AI workflows pharma is racing to deploy. By design, value compounds in this HYFT representation layer, not in any individual model that runs on top of it.
Forward-Looking Statements
This press release contains forward-looking statements within the meaning of applicable United States and Canadian securities laws. Forward-looking statements are based on management’s current expectations and assumptions and are subject to known and unknown risks, uncertainties, and other factors that may cause actual results to differ materially, including: risks relating to patent prosecution and the outcome of examination of European patent application No. EP26187897.9 covering the Company’s high-dimensional vector data structure; the possibility that claims may be narrowed, rejected, or successfully challenged; the scope, issuance, enforceability, and competitive value of any patent that may issue; risks relating to the protection and enforcement of intellectual property; risks that preclinical results described in this release, including the dengue binding-level data and the influenza functional-constraint findings, may not be reproduced and may not translate into neutralization, safety, efficacy, durability, or clinical or commercial success, and that these programs remain preclinical and subject to substantial further study; the size, growth, and addressability of the markets referenced, which are based on third-party estimates that may prove inaccurate; the role of the patented architecture in HYFT® Technology, ReefIQ™, the biological representation, and the LensAI™ platform; the pace and degree of customer and market adoption of the Company’s infrastructure; the technical performance of AI-based discovery methods and of the underlying compute infrastructure; competition; regulatory determinations; and capital markets conditions. Additional information is available in MindWalk’s Annual Report on Form 20-F and other filings on SEDAR+ (sedarplus.ca) and EDGAR (sec.gov/edgar).
Trademarks
HYFT® is a registered trademark of MindWalk Holdings Corp. LensAI™, ReefIQ™, HYFT Base™, HYFT Matrix™, HYFT Prime™, and B cell Llama™ are trademarks of MindWalk Holdings Corp. or its subsidiaries; ReefIQ™ registration is pending. All other trademarks are the property of their respective owners.
References
[1] NVIDIA Developer Blog, “Build an AI Scientist for Life Science Discovery with NVIDIA BioNeMo Agent Toolkit,” developer.nvidia.com.
[2] He J, Lai H, Saigiridharan L, Ghiandoni GM, et al., “Democratising real-world drug discovery through agentic AI,” Drug Discovery Today, 2026; 31(2):104605. DOI: 10.1016/j.drudis.2026.104605 (PMID 41548711). Describes AstraZeneca’s ChatInvent agentic system.
[3] AI in drug discovery market size and forecast: Fortune Business Insights, Artificial Intelligence (AI) in Drug Discovery Market, 2026 (approximately US$5 billion in 2026); Astute Analytica, AI in Drug Discovery Market, February 2026 (more than US$8 billion by 2030). Third-party estimates vary; figures are presented for context only.
[4] Global pharmaceutical research and development spending: Fierce Biotech, The Top 10 Pharma R&D Budgets of 2025; Statista, Worldwide Pharmaceutical R&D Spending. Industry-wide annual R&D investment exceeds US$250 billion.
Contacts
Investor Contact
Louie Toma, CPA, CFA
Managing Director, CoreIR
[email protected]
