The Data Revolution in Drug Safety: De-Risking Pharma with Proteomics

MONTREAL, QC & WASHINGTON, D.C. – June 25, 2026 – The pharmaceutical industry operates on a razor's edge, where the path from discovery to market is a multi-billion dollar gauntlet. A significant contributor to this risk and expense is drug-induced toxicity, a silent saboteur that derails promising compounds, often late in development. Now, a strategic partnership between proteomics innovator Nomic Bio Inc. and the HESI Global OASIS Consortium aims to fundamentally de-risk this process, shifting safety assessment from a reactive hurdle to a predictive science.

The collaboration will leverage Nomic’s high-throughput protein profiling technology to create an unprecedented map of liver toxicity, generating foundational data intended to redefine how we ensure medicines are safe for human use. This isn't merely an academic exercise; it's a direct response to a changing regulatory landscape and a strategic move to build the data infrastructure for the next generation of AI-driven drug development.

A New Mandate for Innovation

The ground for this new approach was tilled by significant regulatory evolution. The landmark FDA Modernization Act 2.0, passed in late 2022, broke with decades of precedent by removing the federal mandate for animal testing in drug development. This legislative shift unlocked the door for so-called New Approach Methodologies (NAMs)—a suite of advanced tools including cell-based assays, organ-on-a-chip systems, and computational models designed to provide more human-relevant data.

While championed for their ethical benefits in reducing animal use, the primary driver for NAM adoption is scientific and economic. Animal models frequently fail to predict human toxicological responses, creating a costly gap between preclinical and clinical results. NAMs promise to bridge this gap by using human cells and biological pathways to generate data that is, by design, more predictive of human outcomes. The challenge, however, has been validating these new methods and generating data on a scale sufficient for regulatory confidence. This is the precise challenge the OASIS Consortium was formed to address, creating a neutral ground for industry, government, and academia to build the tools for a post-animal testing world.

Inside the Data Engine: From Proteins to Predictions

At the heart of this new collaboration is Nomic Bio's proprietary technology. The company's Omni 1000 platform will analyze over 3,000 samples from human liver cell models (HepaRG) treated with 168 different compounds known to have varying effects on the liver. The goal is to capture a detailed snapshot of the proteome—the complete set of functional proteins—as it responds to chemical stress.

Proteins are the functional workhorses of the cell, and their changing levels can signal everything from cellular stress and inflammation to the activation of injury and death pathways. A key technological advantage Nomic brings is its ability to report protein measurements in absolute concentration units (pg/mL). This seemingly technical detail is a strategic game-changer. Unlike relative quantification, which can vary between experiments, absolute units create a universally comparable, analysis-ready dataset. Data from one experiment can be directly and reliably compared to another, a critical feature for building the large, interoperable datasets required for robust machine learning.

“Human-relevant datasets are essential to advancing more predictive safety assessment,” said Milad Dagher, CEO of Nomic. “This study is designed to generate large-scale proteomic datasets across diverse compound exposures, creating resources that can support toxicity prediction efforts while providing pathway-level biological context alongside computational approaches.” This provides not just the raw data for prediction, but the mechanistic context to understand why a prediction is made, a crucial step for both scientific understanding and regulatory trust.

A Consortium Approach: Building the AI-Ready Future

Nomic's contribution is a vital component of the broader vision of the HESI Global OASIS Consortium. OASIS, or Omics for Assessing Signatures for Integrated Safety, is a multi-stakeholder effort designed to integrate various high-dimensional biological data streams to transform chemical safety assessment. Nomic’s proteomics data will not exist in a vacuum; it will be combined with transcriptomics (gene expression) and Cell Painting (high-content imaging) data from the same experiments.

This multi-omics approach creates a dataset of unparalleled richness, providing a holistic view of a cell's response to a compound. It is this very richness and scale that makes it perfect fuel for artificial intelligence. AI and machine learning algorithms excel at identifying subtle patterns in vast, complex datasets that are impossible for humans to discern. The OASIS project is, in effect, creating the ultimate training ground for AI models designed to predict liver toxicity before a compound ever enters a clinical trial.

“The goal of OASIS is to generate high-quality, shared scientific resources that help move safety assessment toward more predictive and human-relevant models,” explained Chrissy Crute, Ph.D., Scientific Program Manager at HESI Global. “The resulting proteomics datasets will be combined with Cell Painting and transcriptomics to support the development of more predictive NAMs while also helping researchers understand the biological mechanisms that drive toxic responses.”

The Public Payoff: De-Risking the Path to Safer Medicines

The long-term strategy extends beyond the immediate partners. A core tenet of the collaboration is the commitment to make the results publishable and the underlying data publicly available. This open-science approach is designed to catalyze the entire field, allowing researchers globally to validate findings, develop new analytical tools, and build upon the foundational dataset without duplicating costly and complex experiments.

For the pharmaceutical industry, the potential payoff is enormous. Drug attrition due to toxicity is a primary driver of cost and failure. By enabling the earlier and more accurate identification of problematic compounds, this initiative promises to significantly reduce the risk profile of drug development. This allows companies to focus resources on candidates with a higher probability of success, accelerating the delivery of safer, more effective medicines to patients. In a world of escalating R&D costs and geopolitical pressures on supply chains, creating a more efficient and predictable pipeline for new medicines is the ultimate competitive advantage.