Manuel Beiran, Ph.D.

Manuel Beiran is a postdoctoral researcher in theoretical neuroscience at Columbia University’s Zuckerman Institute. He received a bachelor’s degree in physics from Universidad Autonoma de Madrid in 2014 and a master’s degree in computational neuroscience from the Bernstein Center for Computational Neuroscience in Berlin in 2016. In 2020, Beiran completed his Ph.D. in computational neuroscience at the Ecole Normale Superieure in Paris, supervised by Srdjan Ostojic. Since 2021, he has been a postdoctoral researcher at Columbia, supervised by Ashok Litwin-Kumar, as well as Kanaka Rajan at the Icahn School of Medicine at Mount Sinai. His research focus is in theoretical neuroscience, specifically in models of recurrent neuronal networks, and the links between connectivity, neural activity and computation. Currently, he is focusing on understanding how to effectively incorporate structural connectivity data into the modeling of neural networks, to suggest new experimental paradigms and test predictions at the neural level.

Principal Investigator: Ashok Litwin-Kumar

Fellow: Anietie Ekanem and Arjun Dibya

Undergraduate Fellow Project: Dynamics of recurrent neural networks with connectome-inspired constraints

In this project, we will develop models of recurrent neural networks whose connectivity between neurons is constrained by the existing data of neural connectivity from the fruitfly brain. We will use mathematical analysis to develop a theoretical framework and carry out numerical simulations using scientific programming to study the relationships between connectivity constraints, network dynamics and neural computation. We will train recurrent neural networks to perform time-varying computations. We will compare networks that are trained using different connectome-constraints, to try to understand the role of specific groups of neurons or synaptic patterns in the connectivity.