research overview
Cancer treatments lead to favorable outcomes only in a subset of patients partly due to significant heterogeneity in tumor cells. Towards developing improved and personalized treatments, the Azizi Lab develops machine learning and AI frameworks to understand the complex cellular ecosystems that drive cancer progression, immune evasion, and resistance to therapy. Our work is motivated by a central question: how do diverse cells within a tumor interact and adapt over time to shape disease trajectories and treatment outcomes?
Tumors are not uniform masses of cells but dynamic, evolving communities. Their behavior emerges from intricate interactions among cancer, immune, and stromal populations that change in space and time. Dissecting these processes requires models that can bridge molecular, cellular, and tissue scales, and learn directly from the richness of patient-derived single-cell and spatial high-dimensional data.
Our laboratory integrates machine learning, artificial intelligence, and statistical modeling with cutting-edge single-cell and spatial, genomics, and imaging technologies to reveal the principles governing tumor ecosystems. We design computational models that capture the temporal and spatial dynamics of the tumor microenvironment, reconstruct regulatory and intercellular communication programs, and identify determinants of therapeutic response and resistance.
Our current efforts focus on developing:
- Integrative spatiotemporal frameworks that decode how multicellular organization and communication networks evolve during cancer progression and metastasis.
- Foundation models for cellular systems that generalize across tissues, modalities, and disease contexts, forming a basis for predictive simulation of tumor evolution and treatment response.
- Causal and generative modeling approaches that uncover the underlying drivers of phenotypic plasticity and immune evasion, and simulate tumor–immune interactions in silico.
By combining data-driven modeling with experimental collaboration and validation, our work aims to transform how we study and treat cancer. We envision a future where AI frameworks can not only learn from patient data but also predict and guide interventions, accelerating discovery and bringing precision medicine closer to reality.
