VSR Seminar with two talks

Start
21st October 2024 11:30 AM
End
21st October 2024 12:30 PM
Location
Jülich Supercomputing Centre, Lecture Hall, building 16.3, room 222

1st talk: Understanding the Hubbard Model using Quantum Monte Carlo

Speaker: Dr. Evan Berkowitz, Theory of the Strong Interactions, IAS-4

Abstract:

The Hubbard Model is a touchstone model of condensed matter physics which describes electrons hopping between sites (ions in a crystal, for example). It can describe strong interactions and yields strongly-correlated electronic dynamics and a rigorous understanding therefore requires a nonperturbative approach. We apply auxiliary-field quantum Monte Carlo and tools from lattice QCD to study these systems.

I will describe how we have leveraged HPC to extract precision physics, focusing on two examples. The first being graphene's quantum phase transition between a semimetal and antiferromagnetic Mott insulator, and the second being state localization on hybrid carbon nanoribbons.

While we have focused on carbon systems, our computational approach is general.

2nd talk: Computational aspects of cryo-electron microscopy

Speaker: Dr. Daniel Mann, Ernst Ruska-Centre for Microscopy and Spectroscopy with Electrons, ER-C

Abstract:

Cryoelectron microscopy (CryoEM) has matured to one of the preferred structural determination methods for big, heterogeneous or flexible biomolecules. At the Ernst Ruska Centre 3 (ER-C-3) we apply and develop high resolution cryoEM with focus on membrane-bound proteins, for example in autophagy research, a central recycling machinery of the cell. In short, thousands of movie snapshots get corrected for motion and optical aberrations, millions of individual protein particles are automatically picked and extracted and subjected to high resolution 3D structure refinement by solving the so-called orientation problem. Further classification in 2D and 3D is applied in order to identify flexibility and increase resolution. JURECA-DC is one of the pillars in our daily work, enabling us to process several Terabytes of raw image data every day through its CPU, GPU and storage resources. This talk will cover essential computational steps that are required to solve high resolution structures from images with examples from autophagy research, ESCRT biology, antibiotic research and CryoEM method development.

Last Modified: 11.10.2024