SUPR
New biophysical models for studying biomolecular interactions
Dnr:

NAISS 2023/5-531

Type:

NAISS Medium Compute

Principal Investigator:

David Van Der Spoel

Affiliation:

Uppsala universitet

Start Date:

2024-01-01

End Date:

2025-01-01

Primary Classification:

10402: Physical Chemistry

Secondary Classification:

10603: Biophysics

Tertiary Classification:

10407: Theoretical Chemistry

Allocation

Abstract

Molecular simulation is an extremely powerful tool that complements theoretical research and experiments. In order to be able to apply molecular simulations routinely, it is imperative to have efficient software that is easy to use, and physical models that are reliable and well-characterized in terms of their predictive power. We are working on a new force field, Alexandria, that is being trained using our soon-to-be-released software, the Alexandria Chemistry Toolkit (ACT), on databases of quantum chemical data. Predictive power is the main design criterion and we will provide quantitative analyses of the quality and robustness of the models. Our main of application area is modelling of amyloid peptides to study the thermodynamics of amyloid fibril formation (Hosseini & Van der Spoel, Protein J. 42 (2023) pp. 192-204). Research into the thermodynamics aspects of peptide aggregation holds the potential to make a tangible impact on our understanding of diseases due to protein misfolding. For instance, we will analyse the effect of known hereditary mutations on the binding strength in aggregates of peptides from amyloid fibrils. A recent review strongly suggests that accurate physical models are needed to study such systems. Based on previous results from the Alexandria force field we are confident that our models will be sufficiently accurate to generate quantitative results for the biomolecular interactions governing fibril formation. Indeed, both method development and applications have so far been very successful in terms of publications, and progress has been discussed in a number of recent review papers. The ACT uses machine-learning for deriving physical force fields from QM databases with a tunable accuracy. This underscores the high significance of further research and development of this software. Both developments and applications would have been possible without the aid of NAISS computer resources. Continued support from NAISS is crucial to progress towards force field with high accuracy.