Molecular dynamics (MD) simulations encompass a range of computational techniques useful to sample the dynamical behaviour of molecules and relate it to their structure. Finding the link between molecular dynamics and molecular function is key to the development of new drugs and, more recently, to the design of enzymes or other macromolecules with a direct use in industry. Nevertheless, understanding molecular dynamics is an extremely complicated task considering the large amount of spatio-temporal information collected by MD simulations, calling for use of machine learning approaches.
In this project you will work alongside researchers in the Department of Chemistry at UoA (Mercadante Lab) to build a machine-learning framework that aims at finding dynamic descriptors of molecular function by analysing the trajectories (atomic positions, velocities and forces over time; provided by Mercadante Lab) on enzymes with well-known structure-function relationship. This will help train models that can readily identify molecule structure conformation and dynamics that mediate certain molecular functions.
Duration and Type
- 12 week summer scholarship in the New Zealand Summer 2021/2022
- At least basic skills in statistics, data mining or machine learning; Intermediate level programming skills in Python.
Supervisor and contact
- Katerina Taskova and Davide Mercadante
- Send CV and transcript by mail to Katerina Taskova