Ayan Paul

About

Ayan Paul is a Research Associate Professor at Northeastern University's Institute for Experiential AI, where he leads the Neural Dynamics Lab. He is also Sponsored Staff at the Channing Division of Network Medicine, Brigham and Women's Hospital, Harvard Medical School. His research program spans computational biology, genomics, protein science, applied AI for biomedicine, theoretical particle physics, and core machine learning. He received his doctorate in theoretical particle physics from the University of Notre Dame, was a postdoctoral fellow at INFN, Rome, and subsequently held positions as a Fellow at DESY, Hamburg, and Senior Scientist at Humboldt Universität zu Berlin before joining Northeastern in 2022.

The Neural Dynamics Lab develops deep learning and classical machine learning models for decoding alternative splicing regulation, protein function prediction and directed evolution, LLM-assisted knowledge aggregation, and computer vision for image superresolution. The lab maintains bioinformatics pipelines for short-read and long-read RNA-seq, MeRIP-seq, eCLIP, and variant calling, and is developing transformer-based large language models for predicting alternative splicing quantification. Current applications include understanding splicing dysregulation in COPD and small cell lung cancer, connecting physical fitness to cognitive ability, and optimizing enzymes for plastic degradation. The lab comprises a team of 20 fully funded members, including research scientists, computational biologists, bioinformaticians, machine learning engineers, data engineers, and graduate students.

Prior to his transition to computational biology, Ayan established a record of contributions in theoretical particle physics. His work on CP violation in charmed meson decays contributed to the theoretical predictions that led to the discovery of CP violation at CERN in 2019. He developed HEPfit, a Markov Chain Monte Carlo-based Bayesian inference framework now widely used for parameter estimation in particle physics, and contributed to the design of future colliders including the FCC, CEPC, and ILC. Together with collaborators at DESY, he introduced interpretable machine learning methods to particle physics phenomenology, with the resulting work published in Nature Reviews Physics.

Ayan has authored over 50 peer-reviewed publications across interpretable machine learning, theoretical particle physics, mathematical epidemiology, computational socioeconomics, and genetics. He has delivered over 30 keynote and invited talks at international conferences and has organized several conferences in machine learning and particle physics. He is also actively involved in technology transfer as the cofounder of CoVis Inc. and KarmaV Inc., and serves on the Steering Committee for the Asian Faculty and Staff affinity group at Northeastern University.

Research Interests

Computational biology, genomics, protein science, applied AI for biomedicine, theoretical particle physics, interpretable machine learning, deep learning, Bayesian inference, and mathematical epidemiology.

Education

Ph.D. in Physics, University of Notre Dame, 2012