Research

Our lab studies lineage plasticity and histologic transformation, through which cancer cells change identity under therapeutic and environmental pressure. Histologic transformation is among the most clinically consequential manifestations of plasticity, allowing tumors to adopt a different lineage and escape therapies directed against their original state. We investigate how these transitions and other forms of molecular mimicry enable cancer evolution, metastasis, and treatment resistance.

We view lineage plasticity as a rewiring of information across DNA, RNA, epigenetic state, and the tumor microenvironment. Using single-cell, spatial, and cell-free genomics, we develop computational methods to quantify discordance among these layers, model changes in cancer state, and identify opportunities to predict or intercept them.

Lung cancer is a major area of biological and clinical expertise in the lab, but similar processes occur in other malignancies. We therefore study genitourinary, gastrointestinal, and hematologic cancers as comparative models, seeking both shared and disease-specific mechanisms of plasticity.

Our research is organized around four related areas.

Xenium spatial imaging and evolutionary trajectory analysis of adeno-to-SCLC transformation

Histologic transformation in solid tumors

We investigate transitions among adenocarcinoma, neuroendocrine, squamous, and other lineage states across thoracic, genitourinary, and gastrointestinal malignancies. These transitions generate extensive tumor diversity, including tumors with features of multiple cancer types, and can undermine therapies directed against a single cell state.

Transformation often proceeds through stem-like intermediate states that can branch toward different fates. We integrate multimodal data to identify the tumor-intrinsic and microenvironmental signals that govern these trajectories and to determine whether intermediate states provide opportunities for therapeutic interception. Comparing transformation across tissues allows discoveries in one cancer type to inform the study of another.

II.

Hematologic malignancies and plasticity across tissue and fluid compartments

Hematologic malignancies are an expanding focus of the lab. We study the progression of indolent leukemia to aggressive lymphoma as a counterpart to histologic transformation in solid tumors, asking whether these diseases use common programs to change identity, escape treatment, and acquire more aggressive behavior.

We also investigate how malignant cells adapt among circulating, floating, and tissue-associated states. These systems, including tissue-forming lymphoma and the adaptation of solid-tumor cells to fluid compartments such as the leptomeninges, allow us to examine how lineage, adhesion, cellular organization, and local environmental signals interact to shape cancer state.

III.

Tumor-immune co-evolution across therapies and microenvironments

Cancer cells evolve within an ecosystem shaped by immune populations, tissue context, and treatment. We study how tumor and immune states change together under these pressures and how their interactions contribute to response, resistance, and relapse.

By comparing tumor and immune remodeling across cancer types, microenvironments, and therapeutic modalities, we seek to identify resistance mechanisms that recur across settings and those that depend on a particular disease or treatment. We also examine how lineage state and tumor heterogeneity influence therapeutic susceptibility and how different treatments redirect subsequent evolutionary trajectories.

IV.

Predicting, detecting, and intercepting cancer-state transitions

We develop machine-learning approaches to predict treatment response and identify emerging resistant states across datasets, disease contexts, and treatment types.

We also connect high-resolution tissue measurements with longitudinal cell-free profiling. Single-cell and spatial genomics resolve the states and interactions within tumors, while cell-free genomic and epigenomic signals provide a potential means of tracking their evolution noninvasively. Our goal is to detect consequential state transitions before they become clinically established and identify opportunities for early intervention.

Our approach

Across these areas, we integrate genomic, transcriptional, epigenetic, spatial, and microenvironmental information from experimental models and patient samples. This enables us to study plasticity at multiple scales (molecular rewiring within individual cells, tumor-immune interactions, disease evolution across treatment and time) and translate that understanding into better ways to classify, monitor, and treat evolving cancers.