Research
We study theoretical aspects of networks and graph machine learning, and develop methods with applications in biological networks and other complex systems.
Graph Neural Networks and Large Language Models
Graph neural networks capture relationships and structural patterns, while large language models interpret text and contextual information. Our goal is to bring these capabilities together.
We aim to develop a hybrid framework that processes graph-structured data alongside text and images, supporting richer reasoning, generation, and inference.
Graph Reconfiguration and Algorithms
Reconfiguration asks whether one valid solution can be transformed into another through a sequence of allowed changes. This perspective offers a way to reason about networks as they evolve.
We plan to study graph reconfiguration problems, analyze their parameterized complexity, and design practical algorithms with applications to robust graph machine learning.
Gene Regulatory Networks
Gene regulatory networks describe how genes and regulatory molecules influence gene expression. Inferring these relationships from noisy, sparse data is a challenging problem.
We aim to develop graph structure learning methods for regulatory network inference and perturbation analysis, including ways to incorporate information from scientific literature.