Biological systems are not uniform. Even within a single diagnosis, patients occupy diverse molecular states that evolve over time. 

My research asks a central question: can we define and model this phenotypic heterogeneity in a way that reveals reproducible structure in complex disease? I develop computational frameworks that integrate multi-omics data, machine learning, and network-based modeling to map how molecular variation organizes into higher-order biological states. Rather than treating heterogeneity as noise, I treat it as a signal, an opportunity to uncover latent disease architectures that transcend conventional clinical categories.

Over the past 15 years, I have worked across oncology, dermatology, pulmonology, and neurodegeneration, collaborating with clinicians and experimental scientists to interrogate transcriptomic, proteomic, and metabolomic data. These projects converge on a unifying objective: identifying molecular drivers and system-level regulatory patterns that explain why seemingly similar patients diverge in their trajectories, responses, or outcomes.

Looking ahead, my program aims to develop scalable computational strategies that link molecular states to clinical phenotypes across scales, from cellular perturbations to patient-level variability. By formalizing heterogeneity as a measurable, modelable property of disease, I seek to contribute to a more mechanistic and predictive framework for precision medicine.

Mentorship is integral to this vision. I am committed to training scientists who are fluent across biological and computational domains, capable of rigorous quantitative reasoning, and comfortable operating at disciplinary interfaces. Building such researchers is, in many ways, as important as building new models.