🎓 Postdoctoral Researcher @ University of Chicago (PME)
⚛️ Computational Chemist • Machine Learning Researcher • HPC Enthusiast
I build methods and tools at the intersection of molecular modeling, machine learning, and high-performance computing. My research focuses on enabling simulations of biomolecular systems at scale, designing enhanced sampling algorithms, and integrating ML potentials into molecular dynamics frameworks.
- Molecular Simulations: OpenMM, GROMACS, ASE, CHARMM-GUI, QM/MM methods
- Enhanced Sampling & Rare Events: Metadynamics, GADES (Generalized Adaptive Dynamics for Escape Sampling), free energy methods
- Machine Learning for Molecular Modeling: Bayesian optimization, VAMPnets, GFlowNets, ML interatomic potentials (e.g., Meta’s UMA)
- Topological Data Analysis: Vietoris–Rips complexes, persistent homology, Betti numbers for hydrogen-bond networks
- High-Performance Computing: SLURM workflows, NSF ACCESS allocations (PI), Dockerized simulation environments
Adapted the saddle-seeking principle of GAD into a practical scheme for enhanced molecular dynamics. Instead of inverting unstable modes, GADES gently damps the softest Hessian direction, allowing trajectories to escape local minima without trapping at saddles.
End-to-end Python framework for automating binding free energy calculations of protein-catalyzed capture agents. Integrates structure building, parametrization, PBMetaD simulations (GROMACS), and post-processing to deliver ΔG and Kd with uncertainty estimates.
A Python implementation of the eigenvector-following transition state search (EF-TSS) algorithm, leveraging the PRFO method and Bofill's Hessian estimation via Gaussian. Users input a settings JSON and initial coordinates, and the tool iteratively converges on a transition state by calling Gaussian for energy and Hessian evaluations, with built-in pytest coverage.
Applied Bayesian optimization to accelerate protein-ligand screening, integrating molecular dynamics with ML-based surrogate models.
- 🌐 Google Scholar
- ✉️ arminshzd[at]gmail.com
⭐️ Always excited to collaborate on projects that combine physics, chemistry, ML, and scalable computing.