XAS calculations for ML training dataset
Dnr:
NAISS 2025/22-878
Type:
NAISS Small Compute
Principal Investigator:
Harald Agelii
Affiliation:
Uppsala universitet
Start Date:
2025-06-10
End Date:
2026-03-01
Primary Classification:
10302: Atom and Molecular Physics and Optics
Webpage:
Allocation
- Centre Storage at NSC: 500 GiB
- Tetralith at NSC: 20 x 1000 core-h/month
Abstract
Resources will be used in a project for developing a machine learning model in predicting X-ray absorbtion spectroscopy (XAS) spectra given a description of the local structure around Cu-atoms in metalloproteins. The computations are required for generating the training dataset for our model. We estimate that we will generate XAS spectra for 5000 different structures from the RSCP protein databank.