I am a biophysicist and data scientist specializing in the modeling of chromatin structure and dynamics. My research integrates stochastic simulations, molecular dynamics, and machine learning, all implemented in Python—my primary programming language and favorite scientific tool.

My research focuses on computational modeling of chromatin architecture, with an emphasis on understanding how the physical principles governing genome organization give rise to the dynamic 3D structure of chromosomes. My work combines multiple complementary approaches, including:

• Molecular Dynamics (MD) simulations coupled with stochastic Monte Carlo models to capture chromatin dynamics across multiple spatial and temporal scales.
• Physics-driven Reinforcement Learning for learning efficient strategies to reconstruct and optimize chromatin conformations.
• Physics-Informed Neural Networks (PINNs) and hybrid machine learning models that integrate biophysical constraints directly into predictive frameworks.

Beyond methodology, I am particularly interested in developing interactive simulation environments where complex chromatin dynamics can be explored intuitively, bringing together scientific visualization, real-time modeling, and concepts inspired by modern game engines. My long-term vision is to make large-scale genome simulations not only physically realistic and computationally efficient, but also highly interactive and accessible for researchers.

So far, I have developed three user-friendly computational pipelines for modeling chromatin structure and dynamics from experimental data, with a strong focus on reproducibility, visualization, and ease of use.

A minimal yet powerful simulator integrating replication dynamics, loop extrusion, and epigenetic spreading.

• Models DNA replication under both physiological and stress conditions.
• Enables exploration of how chromatin structure and epigenetic state influence replication.
• Currently applied to study replication stress in cancer.

“Simplicity is the ultimate sophistication.” — My guiding principle in design and code.

A multiscale simulation platform for modeling chromatin across all levels of organization.

• Supports modeling from nucleosomal arrays to chromosome-scale domains.
• Designed to be user-friendly and extensible for both research and educational purposes.

A stochastic model of loop extrusion capable of reproducing experimental Hi-C heatmaps.

• Captures dynamic interactions mediated by loop extruding factors.
• Useful for testing hypotheses about CTCF-cohesin-mediated chromatin folding.