IB Acquisition details GNQ model's 200M+ causal links
Q-BRM is described as quantum-infused but not quantum-dependent, with production use on classical compute today.
Sentiment and the balance of points
Rhea-AI Sentiment reads the wording of the document, how positive or negative its language is on a 1 to 5 scale. The balance of points shown with the takes weighs what the document actually discloses, so the two can disagree, for example when a trial that missed its main goal is described in upbeat language.
Rhea-AI Filing Summary
IB Acquisition Corp. filed a communication concerning an IBM-issued press release describing a collaboration among GNQ InSilico, IBM Consulting and AWS. The release describes GNQ’s BioAvatar, Q-BRM and BioVerdict products and an IBM-led professional-services offering through AWS Marketplace.
It says Q-BRM’s knowledge graph contains over 200 million causal relationships and describes three reasoning levels: observation, mechanism and causation. The release says the quantum-infused platform is not quantum-dependent and is production-ready on classical compute today, with quantum hardware used for selected molecular binding-energy and therapy-optimization problems. BioAvatar is described as creating biological and patient-specific digital representations; BioVerdict supports traceable outputs, GxP-compliant logging and 21 CFR Part 11 audit trails, with human review before submission.
Insights
Analyzing...
Key Figures
Key Terms
counterfactual simulation technical
sample-based quantum diagonalization (SQD) technical
reaction kinetics technical
GxP-compliant output logging regulatory
21 CFR Part 11 audit trails regulatory
FAQ
AI-generated questions and answers. How Rhea-AI works. Not financial advice.
What is Q-BRM and how large is its knowledge graph?
How can organizations access IBM’s quantum AI for precision medicine?
AI-generated analysis. How Rhea-AI works. Not financial advice.
Filed by IB Acquisition Corp.
pursuant to Rule 425 under the Securities Act of 1933,
as amended and deemed filed pursuant to
Rule 14a-12 under the Securities Exchange Act of 1934,
as amended
Subject Company: IB Acquisition Corp.
(Commission File No. 333- 297710)
Date: September 14, 2026
The following press release was issued by IBM on September 14, 2026, and includes references to GNQ Insilico, Inc. The press release is published at: https://www.ibm.com/think/perspectives/biological-reasoning-life-sciences.
Beyond prediction: Bringing biological reasoning to life sciences R&D with GNQ BioAvatar on AWS
For decades, life sciences organizations have sought to use data and scientific insights to make faster, smarter decisions across the drug development lifecycle. Advances in genomics, imaging, real-world evidence and AI have created more access to data than ever before. However, drug development remains costly, slow and highly uncertain. The challenge is no longer access to data but turning it into actionable insights that improve outcomes.
A key limitation is that many analytical approaches focus on prediction rather than causality. AI can identify patterns, predict outcomes, prioritize targets and identify potential responders but it often struggles to explain why something happens or how an intervention could change the outcome.
This challenge is not a limitation of scale or compute—it is a structural gap. A model trained on associations learns what tends to co-occur, but it has no representation of the mechanisms that produce those associations. When a drug or patient population falls outside the training distribution, the model’s predictions become unreliable in ways that are difficult to detect.
Drug R&D requires more than statistical correlation. Researchers need to understand what drives disease progression, why patients respond differently and what risks might emerge before costly clinical trials. This need requires deeper biological understanding and causal reasoning—moving from “What will happen?” to “Why will it happen?” and “ If we intervene, what happens?”
A shift toward causal intelligence
The next generation of AI in life sciences must go beyond prediction to help researchers understand biological systems, uncover cause-and-effect relationships and evaluate potential interventions.
This vision drives the collaboration between GNQ InSilico, IBM Consulting® and AWS, helping organizations turn biological, clinical and multi-omics data into deeper insights for earlier R&D decisions.
At the center is the GNQ suite of products (BioAvatar, Q-BRM, BioVerdict), developed by GNQ leveraging IBM Consulting and deployed on AWS. BioAvatar creates explainable digital representations of biological systems and patient populations, helping researchers understand how disease mechanisms, biological pathways, patient characteristics and potential treatments interact.
The 3 levels of reasoning in quantum-infused biomedical reasoning model (Q-BRM): From observation to causal understanding
Most AI platforms in life sciences today operate at a single level of reasoning: observation. They identify what tends to co-occur across large datasets, such as which biomarkers correlate with poor response or which compound features associate with toxicity signals. This level of reasoning is useful, but it is not sufficient for the decisions that matter most in drug development.
GNQ’s Q-BRM is designed to reason across three progressively deeper levels.
Level 1: Observation (What)
Pattern recognition across patient populations and biological datasets. Here is where most AI platforms operate. Example: dexamethasone correlates with poor response in a subset of multiple myeloma patients. It is useful as a starting signal, but a prescriber cannot act on correlation alone.
Level 2: Mechanism (How)
Causal modeling of what happens inside a biological pathway when a specific drug intervenes. Example: Q-BRM traces why dexamethasone fails in a specific patient—an HSD11B1 loss-of-function variant in the bone marrow tumor microenvironment prevents cortisone-to-cortisol reactivation at the tumor site. The drug is therapeutically active systemically but inert precisely where it needs to work. This level is where drug target validation, mechanism-of-action reasoning and safety signal explanation become possible.
Level 3: Causation (Why)
Individual-level causal attribution—and counterfactual simulation. Example: Why did this specific patient experience full systemic toxicity with zero anti-tumor benefit? And how would treatment with carfilzomib have changed the outcome? This level is where true patient-level precision medicine operates and at which a regulatory submission can include mechanistic justification rather than statistical association alone.
Moving from level 1 to level 3 is not a technical upgrade; it is a scientific one that requires a causal model of biology, not a larger training dataset. This architectural distinction separates Q-BRM from conventional multi-omics AI platforms.
Understanding biology through causal intelligence
Q-BRM encodes biological knowledge not as a lookup table but as a computational causal model. Its knowledge graph contains over 200 million causal relationships, with edges carrying reaction kinetics derived from established pathway databases, not merely associative weights. This design transforms the graph into a substrate for mechanism-level inference: the platform can reason about what happens to a pathway when a specific gene is perturbed, not just what tends to co-occur with a specific outcome.
It helps researchers answer critical questions such as:
| – | What biological mechanisms drive disease? |
| – | Why do patients respond differently to the same therapy? |
| – | How might changes in a biological pathway affect outcomes? |
| – | Where could potential safety risks emerge? |
| – | Which therapeutic strategies have the strongest biological rationale? |
By combining multimodal data with causal reasoning, Q-BRM helps researchers move beyond prediction to understand the underlying drivers of disease and treatment response. With analytics grounded in auditable biological mechanisms, Q-BRM produces findings that researchers, clinicians and regulators can follow, verify and act on. This capability is the foundation of trusted AI in a regulated environment: not just a prediction, but a causal chain.
Where quantum computing changes the science
Q-BRM is quantum-infused—and it is worth being specific about what that means because quantum computing is not a general accelerant for life sciences AI. It matters precisely where classical computation hits a fundamental wall.
Two such walls exist in causal biological reasoning:
First, molecular binding energies. For a growing class of therapeutically important molecules (including covalent inhibitors, metal-coordinating drugs and novel modalities involving modified nucleotides), the dominant interaction physics are quantum mechanical in nature. Classical simulation cannot accurately model electron correlation, covalent bond formation or spin-state transitions that govern binding affinity and selectivity in these compound classes.
Q-BRM incorporates sample-based quantum diagonalization (SQD), which computes molecular Hamiltonians on quantum hardware. This process provides binding energy calculations from first principles rather than statistical approximation.
These computations are not just theoretical. They enable mechanism-level pharmacology at the molecular level, answering the question “Does this variant change how this drug physically binds to its target?”—a causal question that classical AI approximates rather than solves.
Second, combinatorial therapy optimization. Selecting the optimal multi-drug regimen across thousands of pathway-drug interactions, resistance mechanisms and patient genomic states is a combinatorial optimization problem. At a biologically realistic scale, this problem is computationally intractable for classical solvers. Q-BRM applies optimization on quantum hardware for this class of problem—enabling level 3 counterfactual reasoning at scale: “across this patient’s entire pathway state, what combination of interventions produces the best projected outcome?”
Q-BRM is quantum-infused, not quantum-dependent. These applications are targeted to specific computationally intractable problems, and the platform is production-ready on classical compute today. Quantum acceleration can be applied selectively where the physics of the problem demands it.
Enabling better decisions earlier
Drug development decisions often involve significant investment despite an incomplete understanding of biological risks and therapeutic potential. BioAvatar brings together relevant scientific and clinical evidence to provide a unified view of drug candidates and their potential outcomes.
This approach enables research teams to identify promising candidates, uncover potential risks, validate therapeutic hypotheses and prioritize development opportunities earlier. It also helps reduce uncertainty, improve resource allocation and increase confidence in critical R&D decisions. The AI does not decide. It surfaces causal evidence for researchers and clinicians to act on. In regulated environments, this distinction is not an optional feature—it is required.
Advancing precision medicine
The same intelligence can be applied at the individual patient level. BioAvatar integrates patient-specific genomic, clinical, biomarker and treatment data to create dynamic digital representations of patients.
Researchers and clinicians can use these models to explore potential treatment responses, compare therapeutic options and anticipate possible outcomes. Because each digital twin is built from that patient’s actual biological context (including their genomic variants, pathway state and treatment history), the counterfactual simulations it generates are patient-specific, not population-average.
Together, these capabilities create a progression from understanding disease biology, improving drug development decisions and enabling personalized medicine.
From insight to regulatory-grade evidence
A causal finding is only as valuable as the organization’s ability to act on it in a regulated context. FDA guidance increasingly expects mechanistic rationale in submissions. The agency has signaled that mechanistically justified evidence can reduce the number of clinical trials required to bring a drug to market, provided the causal reasoning is auditable and defensible.
GNQ BioVerdict is built for this environment. All outputs are traceable to specific biological pathway mechanisms. The platform supports GxP-compliant output logging and 21 CFR Part 11 audit trails. By cross-referencing pharmacovigilance signals, clinical evidence and pathway-level mechanistic rationale into submission-ready narratives, agentic AI can accelerate regulatory authoring. Meanwhile, human review gates ensure that no output reaches a submission without qualified sign-off.
The result is AI that does not just generate insights. It is AI that generates evidence packages that regulatory reviewers can follow and act on.
Why AWS?
The potential of BioAvatar, BioVerdict and Q-BRM depends on the ability to securely manage, integrate and analyze massive volumes of biomedical data.
AWS provides the scalable foundation to support the GNQ suite or products, enabling high-performance computing, large-scale data processing, AI workloads and collaboration across research environments. Its security, governance and compliance capabilities also help address the stringent requirements of regulated life sciences organizations.
For organizations seeking to scale AI across complex R&D environments, cloud infrastructure is no longer simply a technology choice—it is a foundational capability for innovation.
Looking ahead
The life sciences industry is shifting from simply generating more data to turning data into deeper biological understanding. R&D is moving toward an AI-native model where every molecule evaluated, experiment conducted and clinical trial generates insights that inform the next decision.
Organizations already have vast amounts of biological, clinical and real-world data. The opportunity now is to use that information more effectively to understand disease, evaluate potential interventions and make better decisions across the molecule-to-patient journey.
The next wave of innovation will come from combining AI with scientific reasoning, mechanistic understanding and simulation-driven research. Through the collaboration between GNQ InSilico, IBM Consulting and AWS, organizations can now access this capability in a production-ready, regulated-environment-compliant form. With this access, researchers can ask better questions, surface hidden mechanisms and ultimately deliver better outcomes for patients.
How can you get started?
IBM, GNQ and AWS have come together to bring the best of this partnership at your fingertips. As a first installment, you can engage with us by leveraging the AWS Marketplace offering IBM’s quantum AI for precision medicine, powered by GNQ Insilico.
This IBM-led professional services offering brings together GNQ’s Q-BRM solution, in tandem with IBM’s industry and integration expertise and AWS’s robust cloud platform, to solve one of the most complex and pressing challenges the life sciences industry is facing.
Whether you are looking for an initial explorative discussion or are ready for a full-fledged implementation, we are ready to help you on your journey.