Research experience

I develop likelihood-free statistical inference methods for copy number variation from single-cell DNA sequencing data. The work combines simulation-based Bayesian approaches, statistical modeling, and machine learning algorithms in R to evaluate inference methods on synthetic and real genomic datasets. This is joint work in progress with the Irving Institute of Cancer Dynamics (IICD). A related book chapter with S. Tavare, "New Uses for Old Stochastic Processes: Models for DNA Pile-Ups from Single-Cell DNA Sequencing," has been accepted and is forthcoming in Poisson Distribution - Methods, Models and Case Studies.

Chimeric antigen receptor (CAR) T-cell therapy has shown much promise in liquid tumors but often fails in solid tumors. This work uses a computational model to examine under what conditions this therapy might fail or be successful. The model includes interactions between cancer cells, CAR T-cells (treatment), and vascular cells (that feed and support tumor growth). From our results, we determined specific tumor conditions in which CAR T-cell therapy is predicted to fail and suggest a combination treatment that might improve the efficacy of the treatment. Namely, we explore two forms of antigen expression in tumor cells: binary, and heterogeneous.

As a part of my undergraduate senior project, I implemented an oxygen diffusion module within an existing 3D agent-based model of breast cancer developed in Dr. Norton's lab. The module includes 2D and 3D partial differential equations (PDEs) and their numerical evaluations using the finite difference method. The vascular diffusion process occurred in four steps: 2D point source diffusion, 2D line source diffusion, 3D cubic patch diffusion, and vascular diffusion of oxygen. Previous research conducted in the lab used a uniform distribution of oxygen in tumor-associated vasculature. This work added a more realistic representation of oxygen concentrations in blood vessels, which will hopefully yield more realistic results.

The project aimed to use graph representations to produce simplified sentences from given data. We implemented a sentence fusion graph, extracted linguistic features from text data, and trained binary classification models to rank the generated sentences. This work was funded by Bard Summer Research Institute (BSRI).