AI-Driven Systems Biology: Transforming Biomedical Discovery and Future Medicine

Our lab conducts rigorous translational research at the intersection of artificial intelligence (AI), systems biology, systems pharmacology, and individualized systems medicine. We develop and apply innovative network-based and AI-driven methodologies to uncover the fundamental mechanisms underlying drug action, drug response, and therapeutic outcomes at the individual patient level.

We are committed to advancing the next generation of personalized medicine by integrating large-scale biomedical and multi-omics datasets to construct comprehensive, patient-specific molecular networks. Through these efforts, we aim to elucidate the pathophysiological roots of complex diseases, enable mechanism-based therapeutic discovery, and inform precision treatment strategies tailored to each individual's molecular profile.

Here, we share how the team's work is paving the way for breakthroughs that can make a real difference for patients and families.

Turning Complexity into Opportunity: Diseases are extraordinarily complex. Genes interact with one another in ways that vary from person to person, influencing how illnesses develop and how patients respond to treatment. Traditional tools often fall short in capturing this complexity. Rather than seeing this as a barrier, my team embraces it. We view complexity as an opportunity, one that inspires us to innovate and discover solutions that can directly benefit patients. Breakthroughs in medicine often begin with a new way of looking at a problem. Over the past decade, my lab has developed powerful scientific frameworks and AI technologies that do just that.

New Ways of Understanding Disease: One cornerstone of our work is a concept called the Gene Utility Model, which asks a simple but important question: How does a gene actually contribute to a disease? Rather than focusing only on mutations, we look at how genes function within the larger network of biological interactions. Using an innovative tool we created, called NetDecoder, we have identified key genes that drive disease, many of which represent promising new therapeutic targets. This approach gives us a much clearer picture of what really matters inside a diseased cell.

Opening the AI "Black Box": AI has enormous potential to transform medicine, but only if we can understand what these models are learning. To address this, we developed a novel algorithm called weight engineering, which allows us to peer inside deep learning systems and see the biological relationships they are uncovering. This gives researchers and clinicians actionable insights, helping us identify new disease drivers that would otherwise remain hidden.

Mapping How Disease Takes Shape in the Body: Another breakthrough from my lab is SPIN-AI, a technology that uncovers how cells are physically organized within tissues spatially. Our discovery that certain genes encode "positional information" for cells opens entirely new possibilities. Spatial organization shapes how tumors grow, how immune cells respond, and how diseases progress. By understanding this internal map, we can better target therapies and intervene earlier in the disease process.

AI That Can Test Scientific Ideas: One of our most visionary inventions is Hypothesis-Driven AI (HD-AI). Traditional AI predicts outcomes; HD-AI can actually test scientific hypotheses. This creates a powerful partnership between human experts and AI, accelerating discovery, refining ideas, and reducing the time it takes to get from insight to clinical impact. This patented approach represents a shift in what artificial intelligence can do for medicine.

Reimagining Clinical Trials: Clinical trials are essential for bringing new treatments to patients, but many promising therapies fail simply because the right patients were not enrolled. To address this, we designed a conceptual framework called Artificial Clinic Intelligence (ACI). ACI can simulate clinical trials, create synthetic patient populations, and highlight which clinical features matter most for treatment success. This approach has the potential to lower costs, improve trial efficiency, and help clinicians match treatments to the patients who will benefit most.

Training the Next Generation and Building a Collaborative Future: Our work thrives in the collaborative, forward-looking environment of the scientific community. We partner with researchers across institutions and around the world, and we are deeply committed to training future scientists, from high school students to graduate students, MD, PhD and MD-PhD trainees. Their energy and creativity fuel our progress and ensure that our discoveries will continue to expand long into the future.

Poised for Progress: Building on Our Momentum:
Scaling Data-Driven Discovery: Expanding our AI and systems biology platforms to reveal deeper disease mechanisms and accelerate target discovery.
Advancing Toward Clinical Trials: Progressing our lead therapeutic and diagnostic programs into IND-enabling studies.
Strengthening a Comprehensive Cross-Disciplinary Collaboration Framework: Integrating all essential disciplines from bedside to bench to byte and back again to drive innovations from clinical need through discovery, translation, implementation, and patient impact.
Growing Impact Across the Clinic: Transforming lab innovations into practical tools and interventions that meaningfully improve care pathways across diverse clinical specialties.

A Vision for the Future of AI-Enabled, Data-Driven Medicine: We are entering a new era in which biomedicine, computation, and clinical care are converging, it's not just a conceptual shift, it's reshaping how discovery, diagnosis, and treatment are done across the entire biomedical ecosystem. Our vision is to harness this convergence to fundamentally change how disease is understood, diagnosed, and treated. Every discovery we pursue, whether built on multi-omics, advanced imaging, digital biomarkers, or next-generation AI models, serves a single, unifying goal: to improve patient care and outcomes. Across our laboratories, data platforms, and clinical environments, we are building an integrated ecosystem where multi-scale biomedical data and human-centered artificial intelligence work together. By decoding the molecular architecture of disease and transforming these insights into clinically actionable tools, we aim to deliver care that is earlier, more precise, and more predictive than ever before. We envision a future where patients benefit from models that learn continuously, diagnostics that anticipate disease before symptoms appear, therapies tailored to individual biology, and clinical trials that adapt in real time. Innovation becomes not just a scientific endeavor, but a direct pathway to better health.