My postdoctoral work with Robert McDougal and Willam Lytton has focused on developing and applying computational neuroscience simulation tools. I have primarily worked on the NEURON simulation platform, with particular focus on the reaction-diffusion module (rxd). I developed a volume-averaged method for reaction-diffusion simulations in the extracellular space and assisted in the improved support for 3D intracellular modeling, both the voxelization and numerical integration. I also improved the performance and scalability, added Just-In-Time compilation, multithreading, and support on multiprocessor parallelization.
I have utilized these improvements in NEURON for modeling ischemic stroke and spreading depolarization with collaborators at SUNY Downstate. These models include greater biological detail, achieved by exploring the parameter space for a regime that could maintain ion homeostasis under physiological conditions and explicitly including the spatial aspects, such as the dendritic arbor and the extracellular space. I have also taken advantage of the improvements to the intracellular simulation to model the spread of PKMĪ¶, to validate a mathematical model of erroneous long-lasting, long-term potentiation due to crosstalk between nearby spines.
I’ve also recently been working on an informatics project to build a machine learning tool to classify papers that contain a COVID-19 immune signature (the key aspects that define an immune response). I am extending this work using natural language processing (SciBERT with PyTorch) to extract key components of immune signatures from research articles to improve curation.
I worked on mathematical and computational models of microelectrode biosensors at Warwick Systems Biology Centre. I completed my PhD at Warwick University’s Complexity Science Doctoral Training Centre. My thesis focused on developing a macroscopic model of adenosine clearance from the available literature and validating it against experiments.