image: Induced pluripotent stem cells in the shape of a heart.
Credit: Valentina Fossati, NYSCF
Developing new drugs is one of the riskiest and most expensive endeavors in science. Today, The Jackson Laboratory (JAX) announced a bold initiative to improve that trajectory, powered by an up to $30 million contract from the Advanced Research Projects Agency for Health (ARPA-H) Computational ADME-Tox and Physiology Analysis for Safer Therapeutics (CATALYST) program. CATALYST is led by ARPA-H Health Science Futures Mission Office Acting Deputy Director Andy Kilianski, Ph.D.
The initiative, called CARDIOVERSE, will leverage mice and human stem cells that model the range of genetic backgrounds found in people, combined with advanced artificial intelligence to create computational “digital twins” of the human heart. By building virtual, beating hearts, JAX scientists aim to dramatically reduce the need for large-animal studies, streamline FDA approval, and ensure safer, faster delivery of new therapies to patients.
“This is a moonshot vision,” said Matt Mahoney, JAX computational biologist and project lead. “Imagine a future where computational models predict drug safety so reliably that it becomes ethical to move forward on computational evidence alone. That would revolutionize drug development, making it far more affordable and accessible.”
Beating heart toxicity at its core
Despite passing early safety standards, more than 90 percent of drugs fail – often due to unforeseen toxicities that emerge late in clinical trials. CARDIOVERSE tackles one of the most persistent and consequential challenges in drug development: cardiotoxicity, when potential drugs can interfere with heart function. It’s a leading cause of clinical trial failure and responsible for up to 15 percent of withdrawals of new drugs after FDA approval. Individuals may experience different adverse effects, which further complicates the approval process.
To address this, Mahoney’s team will train AI algorithms using data from mice and human cells that replicate human heart function in a dish and capture the genetic variability present across patient populations. JAX is a world leader in mouse studies and the use of human stem cells, and with its recent acquisition of the New York Stem Cell Foundation (NYSCF), the team will profile these cellular models at an unprecedented scale using cutting-edge robotic automation.
This data will power AI models that go beyond safety predictions. They will uncover how genetic differences control individual responses to treatment, offering a more complete picture of cardiotoxicity across patient populations. The result: advanced AI models capable of predicting drug safety across a broad spectrum of human genetic makeups.
“Too many promising medicines never reach patients because we can’t predict early enough who they will help and who they might harm,” said Lon Cardon, JAX president and CEO. “With CARDIOVERSE, we’re building virtual hearts that more accurately reflect the broad genetic backgrounds of real people. This means we can identify which drugs are safe and effective for certain patients, and which may not be, so every person has a better chance of receiving the treatment that’s right for them.”
Working with partners from the University of Michigan, InSilicoTrials Technologies, and the University of Connecticut Health Center, Mahoney’s team will also investigate the molecular mechanisms inside the heart. They will study gene activity and metabolic changes following drug exposure—insights that could reveal biomarkers and genetic risk factors that make certain populations more vulnerable to toxicity.
Building on this foundation, the CARDIOVERSE team will assemble one of the largest data resources of its kind. Using mice and human cells that represent a broad range of genetic profiles, they will generate the critical data needed to train AI models, creating a powerful tool to make cardiotoxicity predictions more accurate than ever before.
“Our AI model will show a population of virtual hearts, each one beating and responding differently to the same drug,” Mahoney said. “That range of responses lets us capture cardiotoxicity outcomes across wide patient groups, identifying what fraction of the population may be at risk and uncovering potential biomarkers for that risk.”
This knowledge could improve patient stratification in clinical trials and give companies greater confidence in advancing drugs to later stages, enabling safer, more precise decisions about which therapies should move forward.
Capturing rare but dangerous reactions
For heart drugs, rare toxicities are a leading reason for failure. Even if only one in a thousand patients experiences a severe reaction during clinical trials or after a drug has been approved, regulators must take that risk seriously.
Current computational models estimate the overall chance a drug will be deemed unsafe, but not who might be affected or how severe the response could be. “It’s like a weather forecast that predicts a 50 percent chance of rain but doesn’t say whether it will be a drizzle or a hurricane,” Mahoney explained.
By simulating hundreds or thousands of virtual hearts across different genetic backgrounds, CARDIOVERSE will identify those rare but serious reactions before human trials begin.
“It will be a big win if we can surrogate large-animal studies with a state-of-the-art computational platform,” Mahoney said. “That would take an expensive experiment off the table and allow more drug candidates from lean startups that are not well capitalized to get to human trials. CARDIOVERSE could help make that happen.”
Project co-leads include JAX scientists Nadia Rosenthal, recognized for her expertise in high-resolution imaging of tiny mouse hearts; Paul Robson, a stem cell biology expert and a lead developer of the genetically varied human iPSC panel; and Travis Hinson, a cardiologist whose lab investigates inherited conditions leading to heart failure. They are joined by Daniel Paull of NYSCF, a pioneer in high-throughput automation for generating iPSCs at scale and overseer of the Global Stem Cell Array®.
About The Jackson Laboratory
The Jackson Laboratory (JAX) is an independent, nonprofit biomedical research institution with a National Cancer Institute-designated Cancer Center. JAX leverages a unique combination of research, education, and resources to achieve its bold mission: to discover precise genomic solutions for disease and empower the global biomedical community in the shared quest to improve human health. Established in Bar Harbor, Maine in 1929, JAX is a global organization with nearly 3,000 employees worldwide and campuses and facilities in Maine, Connecticut, California, Florida, New York, and Japan. For more information, please visit www.jax.org.