Valinor Discovery Unveils AI-Powered Platform to Simulate Drug Response Before Clinical Trials

Salt Lake City, Utah - May 14, 2025

Salt Lake City-based Valinor Discovery has emerged from stealth, announcing its mission to radically reshape clinical drug development using machine learning-powered simulation. With a bold vision to automate drug development and reduce failure rates, Valinor is developing clinically translatable models trained on matched primary cell and assay datasets—enabling virtual patient simulations that predict therapeutic efficacy before the first patient is dosed.

"Clinical drug development is failing us,” said Josh Pacini, Co-Founder and CEO of Valinor. “Despite consuming more than 70% of pharma R&D budgets, 90% of drugs still fail in clinical trials. Humanity is still waiting for effective treatments for devastating diseases like cancer and Alzheimer’s. We believe there’s a better way."

Valinor is addressing the critical gap in clinical data by pioneering the collection of rich, longitudinal datasets—integrating both primary cells and clinical assay data from the same patients. By capturing matched data ranging from cellular samples to patient biopsies, Valinor creates virtual patient models that simulate drug performance in real-world scenarios—not just in vitro.

“Our mission is to fully automate clinical drug development by accurately predicting therapeutic efficacy long before a patient is dosed,” added Pacini. “By simulating treatment responses virtually, we empower drug developers to shift from risky guesswork to precise, data-driven science.”

Zhanel Nugmanova, Co-Founder and Chief Scientific Officer, explained: “By linking molecular perturbations to actual clinical assay data, drug hunters can test therapies virtually in a matter of weeks before investing years at the bench or in the clinic.”

Strategic Collaborations & Platform Access

Valinor is forming strategic partnerships to extend the reach and rigor of its platform.

• Latch Bio Collaboration: In partnership with Latch Bio, Valinor offers a compliant, no-code interface enabling researchers to apply its models quickly and intuitively. Users can deploy ready-to-use workflows for core processes like patient stratification, hit-to-lead analysis, and biomarker panel selection—either out of the box or tailored to specific disease indications.
• Helmholtz Munich: Valinor is collaborating with the Theis Lab at the Computational Health Center at Helmholtz Munich to set new industry benchmarks for drug perturbation prediction. “Valinor’s approach represents an exciting step forward in predictive drug development,” said Dr. Fabian Theis. “We’re excited to work with them to create more robust, clinically translatable benchmarks, which we will add to the OpenProblems platform.”

• Stanford Medicine: Valinor is also partnering with the Montgomery Lab at Stanford, led by Dr. Stephen Montgomery, a key contributor to the GTEx project. The collaboration will apply Valinor’s perturbation models to evaluate new therapeutic compounds for Alzheimer’s Disease.

Importantly, Valinor does not develop its own therapeutics. "Everything we build is designed to power our pharmaceutical customers," stated Pacini. The company’s goal is to equip biopharma partners with tools to bring more effective drugs to market faster and at lower cost.

Scientific Advisors

Valinor has assembled a high-caliber advisory board representing expertise across machine learning, clinical operations, and drug discovery:

• Stephen Montgomery, PhD – Stanford University
• Chase Neumann, PhD – Associate Director of Oncology, Recursion
• Matt Donne, PhD – Former Head of Operations, Spring Science
• Bryan Norman, PhD – Former SVP of Lead Generation Chemistry, Enveda Biosciences; Senior Researcher, Eli Lilly
• Tim Sullivan, PhD – Chief Business Officer, Infinimmune

Valinor Discovery is building a new foundation for therapeutic R&D by combining generative machine learning with patient-matched, longitudinal multi-omics data. Its platform enables accurate simulation of transcriptomic changes, protein levels, methylation patterns, and clinical assay outcomes—prior to patient dosing. Valinor’s models aim to de-risk development, reduce costs, and accelerate breakthroughs in oncology and beyond. Its entire team is based in Utah.

To learn more to inquire about custom models for drug candidates, visit www.ValinorDiscovery.com, view company video below, or view Investing 101 video of investor and podcaster, Kyle Harrison, interviewing Valinor Co-Founder and CEO, Josh Pacini.

Automating Clinical Trials (with Josh Pacini @ Valinor)

“Biology is not an indecipherable black box; it’s living code that we can model and understand.”

Investing 101Kyle Harrison