Research

Modeling Intracellular regulation

The goal of this research area is to elucidate how complex spatio-temporal interactions between signaling molecules (e.g. Rho GTPases), cytoskeleton remodeling, and the extracellular matrix (ECM) modulate cells’ morphology and migration phenotype. Toward this end, we develop new models and mathematical methodologies to assess how the multi-layered, non-linear, feedback interactions (Panel a) between these processes contribute to cellular dynamics. One key outcome of my work was to identify how interactions between Rho GTPase and the ECM, along with key molecular factors influence the persistence of melanoma cell migration (Panel b), which is known to be altered in more aggressive cancer cells, potentially contributing to metastasis. My work has also yielded fundamental new insights about how feedback between signaling and actin remodeling generate the sub-cellular waves of actin activity found in many cell types. I posited (later confirmed experimentally) a novel type of “conservative excitable system” that explains the genesis of both polarity and the formation of these dynamic intracellular waves (Panel c), which had previously been attributed to different underlying mechanisms.

Insulin regulation and anomalous particle dynamics

I have worked with Irina Kaverina (VU) to investigate the influence of cytoskeletal transport and anomalous motion on insulin dynamics in Pancreatic Beta cells (Panel a). Initial results (Panel b) have shown that microtubules have a critical role in negatively regulating insulin secretion and that that mechanism is in overdrive in diabetic mouse cells. This work has led to a more foundational, theoretical investigation of how anomalous motions of cellular particles influence their localization. Initial results (in preparation) suggest that one of the most commonly posited explanations for sub-diffusion (so called Generalized Langevin dynamics) leads to the genesis of depleted boundary layers (Panel c). If these were present in cells, it would be of profound importance. If not, it will point to a serious flaw in our understanding of the physics underlying anomalous particle dynamics.

Development of Multicellular structures

A central question of developmental biology is how the properties of individual cells and interactions between them give rise to organization? We develop multi-scale and stochastic models capable of integrating gene regulatory, biochemical, and mechanical interactions at the cellular level to assess how they influence morphogenesis. Our initial investigations (with Ken Cho, UCI) in this area demonstrated that gene regulatory noise is effective at and potentially necessary for robust and reproducible organization of the early developing embryo. I am continuing this line of research to work with an interdisciplinary team of researchers at Vanderbilt to study how 3D epithelial structures such as cysts develop (Panel c), and elucidate how cellular properties influence their organization. This work also demonstrated, for the first time, that structural features of stochastic gene regulation, such as the asymmetry in noise levels (Panel b), may be of functional importance (Panel a). We are thus also working to understand the effects of structural characteristics of gene expression noise, which are usually overlooked, on the so-called epigenetic landscape.

Human decision-making

One of the central goals of cognitive neuroscience is to couple theoretical models with data to probe the unobservable mechanisms governing decisions. We developed and refined a cutting edge Bayesian methodology for fitting complex models to data, which radically extends the class of models / theories that can be quantitatively tested experimentally. We have since used these methods to quantitatively explore questions that were primarily investigated qualitatively, such as how people make decisions when the information available to them changes and how decisions processes can be optimized. As an extension of this work, we have recently formed a collaboration with the VUMC Dept. of Pathology to investigate how various cognitive factors influence image based diagnostic decisions by physicians and medical technologists.