James Berger Ph.D.At the Kimmel Cancer Center, researchers are revolutionizing cancer drug discovery with the power of artificial intelligence (AI), significantly speeding up the process while reducing costs. Dr. James Berger and his team envision a future where identifying effective therapies takes just years instead of decades, all while ensuring precision targeting of cancer cells. By harnessing vast genetic datasets and employing innovative AI tools, they aim to match specific patients with the right treatment, ultimately accelerating approvals and enhancing patient outcomes. As AI reshapes the landscape of drug development—from small molecules to biologics—this groundbreaking approach promises to deliver new and improved cancer therapies faster and more affordably than ever before. With the right investments and talent, Dr. Berger believes a new era of smarter, patient-focused cancer treatment is on the horizon.
AI and Cancer Drug Discovery
Drug discovery is among the most complex and costly types of cancer research. Now, however, with an assist from AI technologies, Kimmel Cancer Center researchers are reimagining cancer drug research and development, launching a new model that will make it faster, more precise, and less expensive to bring new cancer medicines to patients.
Traditionally, finding a new cancer drug could take more than a decade and cost more than a billion dollars,” says James Berger, Ph.D., the Michael and Ann Hankin and Partners of Brown Advisory Professor in Scientific Innovation and Co-Director of the Cancer Chemical and Structural Biology Program. “With AI, we see the potential to substantially compress both the time and cost of the process.
From Broad Shots to Precision Targets
Fifty years ago, cancer therapy was largely trial and error. Chemists tested chemicals that killed cells, hoping they destroyed cancer faster than healthy tissue. It was a blunt tool, and side effects, from hair loss, nausea and fatigue, and suppression of immune cells were harsh for patients.
Today, precision medicine is the goal. We have new therapeutics in the arsenal, including immunotherapies and gene-targeted medicines. Scientists increasingly know more about how cancers are driven by genetic mutations and molecular pathways, and new drugs aim to hit those targets with surgical accuracy or employ immune cells to destroy cancer cells. The progress has been revolutionary, but finding the right patient-drug match remains a challenge.
This is where AI could excel, Berger explains. By analyzing mountains of genetic data, AI is poised to help researchers identify which patients are most likely to benefit from a particular therapy. Our researchers may soon be able to design clinical trials that better match specific drugs to specific patients. The result, he says, should be faster approvals, lower costs, and treatments that reach patients sooner.
At the Kimmel Cancer Center, this represents a chance to bring its long history of excellence in drug discovery into a new era.
Smarter, Faster, Cheaper
AI is already helping Dr. Berger and colleagues make inroads into nearly every type of therapy, from small molecules and peptides to RNA-based medicines and biologics.
Dr. Berger’s expertise is in small molecules, cancer medicines created from tiny chemical compounds that can easily enter cells and interfere with proteins and protein pathways that fuel cancer growth.
He collaborates with scientists working in the other areas of drug development. Jun Liu, Ph.D., who co-directs the Cancer Chemical and Structural Biology Program with him collaborates on small-molecule drugs but is also using AI to help develop peptide drugs, therapeutics made from the building blocks of proteins that work by mimicking or directing drugs to cancer cells and also by stimulating immune responses. Jeff Coller, Ph.D., Bloomberg Distinguished Professor of RNA Biology and Therapeutics, is exploring how AI can accelerate the development of RNA-based medicines, genetic messengers that edit the instruction manual and change how cells make proteins to silence genes that support cancer growth, replace missing instructions, and train the immune system to respond to cancer cells. Biomedical engineers Jamie Spangler, Ph.D., and Jeffrey Gray, Ph.D., are focused on biologics – in particular, antibody therapies – which are made from living cells or lab-made versions of natural proteins and which typically work on the outside of cancer cells to make them visible to the immune system, block growth, or deliver anticancer drugs directly to cancer cells.
AI is helping these scientists predict and analyze protein structures, search vast chemical libraries, and even design new drug candidates with extraordinary speed and accuracy.
In the past, scientists might have biochemically screened tens of millions of compounds in hopes of stumbling on one that worked. AI changes the approach. By predicting the natural 3D shapes proteins take and how they interact with other molecules, researchers can now design a handful of promising drug candidates on a computer and move them straight into laboratory testing. Pathologists Laura Wood, M.D., Ph.D., and Ralph Hruban, M.D., are already doing this kind of work in pancreatic cancer.
The promise is not only in creating new drugs but also in improving old ones. By using AI to identify why cancers develop resistance, existing therapies can be re-engineered to work better and longer.
Despite the power of these tools, Dr. Berger emphasizes that involvement and oversight of scientists and clinicians remain essential to drug discovery.
AI can propose molecules, but it can’t yet account for every nuance of efficacy, safety and biology,” he says. “We still need experienced researchers across a range of life science and clinical disciplines to keep us on track.
A New Vision
Johns Hopkins already has world-class basic scientists and clinical researchers working together. With AI, Dr. Berger sees a key opportunity to enhance this expertise. His vision includes AI experts who specialize in drug design as members of the School of Medicine’s research team. In the exploding field of AI, these new collaborators—which include specialists in computer-guided chemistry and biology, protein and nucleic acid engineers, and antibody designers—are in short supply and high demand, but Dr. Berger believes they are essential to making real progress.
Ultimately, it will save money and lives, but on the front end, adding these experts and the technology required to do the work requires a multimillion dollar investment, ranging from $10 million to $25 million to build the infrastructure and recruit senior experts.
“We have the scientists who know the targets inside and out. Sometimes they’ve spent decades studying them,” says Dr. Berger. “What we need now are the AI researchers and the tools to match that expertise.
Working together, he believes they can make drug discovery faster, smarter, and more patient focused. From designing antibodies in weeks instead of years, to reimagining small molecules for hard-to-treat cancers, AI is poised to accelerate progress against cancer in ways once thought impossible.
The vision is ambitious but achievable. In the next five to 10 years—perhaps sooner—Dr. Berger believes small groups of Kimmel Cancer Center scientists could design a new cancer drug with just a handful of people, using off-the-shelf computational tools. The time and cost savings could be staggering. What once took a dozen years and over a billion dollars could shrink to 3-4 years and many millions less, he says.
Academic research thrives on fresh ideas and the freedom to take risks,” Dr. Berger says. “Instead of thousands of blind shots, AI gives us more targeted shots on goal and the chance to succeed faster for patients.