Valinor Discovery Launches to Simulate Drug Efficacy in Virtual Patients
Valinor Discovery, a San Francisco-based startup, has launched from stealth with a focus on simulating how patients might respond to therapies—before any clinical testing begins. The company is developing generative machine learning models trained on matched multi-omics and clinical assay data from individual patients, aiming to support drug developers in evaluating compounds, identifying biomarkers, and stratifying patient populations.
The company is initially focused on oncology, using its models to simulate treatment response in cancer-related clinical settings.
To train these models, Valinor is generating proprietary datasets that integrate transcriptomic, proteomic, and methylation profiles with clinical assay results from the same patients. The company states that these simulations are intended to predict transcriptomic changes, protein abundance, methylation status, and clinical assay outcomes—supporting early-stage decisions like compound prioritization, biomarker identification, and patient stratification.
In a LinkedIn post, CEO Joshua Pacini points to "the scarcity of high-quality clinical data significantly limits the development of translatable biological models". He wrote that Valinor is addressing this by developing longitudinal datasets that combine primary cell measurements with clinical assay data from the same patients. According to Pacini, this matched data—from cellular samples to biopsies—will be used to build virtual patient models capable of simulating drug performance in patients rather than in isolated cells.
Valinor is also working with academic groups to evaluate its models and help define standards for perturbation-based predictions across therapeutic areas.
Valinor has announced a collaboration with Prof. Fabian Theis and the Computational Health Center at Helmholtz Munich to develop benchmarking standards for perturbation models. The benchmarks are intended to support evaluation across therapeutic areas and drug types and will be shared via the OpenProblems platform.
A separate partnership with the Montgomery Lab at Stanford Medicine will focus on applying Valinor’s models to the evaluation of compounds targeting Alzheimer’s disease. The lab, led by Prof. Stephen Montgomery, has contributed to large-scale transcriptomic studies, including the Genotype-Tissue Expression (GTEx) project—a public resource that analyzed tissue samples from nearly 1,000 people to understand how genetic differences affect gene activity across organs.
Platform Access and Integration
Valinor is offering its platform through a hosted interface developed with Latch Bio. The platform provides no-code access to models and workflows designed for use cases such as hit-to-lead analysis and biomarker panel selection. The company states that its services are intended for use by biopharmaceutical developers and that it does not develop its own therapeutics.
Valinor is also collaborating with -omics sequencing companies to develop custom models trained on data from biopharma-led clinical trials, both ongoing and historical.
The company’s advisors include individuals from both academic and industry backgrounds:
- Stephen Montgomery, Stanford University
- Chase Neumann, Recursion
- Bryan Norman, formerly Eli Lilly and Enveda Biosciences
- Tim Sullivan, Infinimmune
- Matt Donne, formerly Spring Science
Topics: Startups & Deals
