Matteo Castelli

PhD

Postdoc

Ab initio Molecular Dynamics (AIMD) ABFE AMBER Bash CHARMM-GUI classical molecular dynamics (AMOEBA) Computational chemistry Computational Drug Discovery Drug Design free energy perturbation GNU/Linux GROMACS High Performance Computing (HPC) integrative structural modeling Markov state models MDAnalysis MDTraj MODELLER Molecular Docking Molecular Dynamics Simulations NAMD Pymol Python RBFE RDKit Schrodinger Maestro SLURM SLURM-HPC Structural Bioinformatics Structure Modelling Structure-Activity Relationship Structure-Based Drug Design Tcl Virtual Screening VMD

Preprint(s) 1st author preprint(s) Peer-reviewed publication(s) 1st author peer-reviewed publication(s) Conference Poster(s) Conference Talks(s) Patent(s) Reviewed Publication(s)

I led the computational drug-design effort in an NIH antiviral consortium targeting Ebola virus entry. Starting from experimental and predicted structures of the viral entry glycoprotein, I built and validated structural models of the target and ran structure-based virtual screening, docking, and pharmacophore analyses. I used molecular-dynamics-based free-energy calculations (relative and absolute binding free energies) to prioritize candidate inhibitors, and worked closely with medicinal and experimental chemists to translate predictions into synthesizable, testable molecules. The campaign advanced compounds to in vivo testing and contributed to a patent application.

I led the construction and simulation of a virion-scale molecular-dynamics model of an HIV-1 particle comprising roughly 250 million atoms, run on a leadership-class exascale supercomputer. This required integrating heterogeneous structural data into a coherent all-atom model, building and equilibrating an enormous membrane–capsid–glycoprotein system, and managing large-scale HPC workflows and trajectory analysis. The work pushed the practical limits of biomolecular simulation scale and produced a physically realistic model of a whole enveloped virion.

I built a fully glycosylated, all-atom structural model of a viral entry glycoprotein by integrating cryo-EM, X-ray, and complementary experimental data, then characterized its conformational dynamics and the role of its glycan shield through extensive molecular-dynamics simulation. This connected sparse, heterogeneous structural data into a complete and dynamic picture of an entry protein directly relevant to antiviral and vaccine design.