Fast and Flexible Nested Sampling via Collective Move Dynamics for Molecular Systems

Method development: NS is a powerful technique to sample the configuration space of even non-spherically symmetric particles. However, efficient sampling of complex molecular systems requires sampling of internal degrees of freedom, and advanced schemes such as configurational bias, bond interchange and cluster move MC.

The project aims to understand their statistical efficiency in the context of exhaustive sampling and incorporate them into NS. Software development: The project will develop a flexible code to implement the advanced MC moves. This will be interfaced to a simulation engine for calculations of energy, and to the NS implementation PyMatNest.

Applications: The new technique will be tested first using a small, flexible molecule, then extending into longer-chained ones, either employing atomistic or coarse-grained 'bead and spring' models. Machine learned potential may also prove tractible-NS is potentially the ideal technique to also generate training data for such models.

The outcomes of the project are three-fold:

1. Produce robust and sustainable software that enables the exhaustive sampling of the configuration space of a wide range molecular substances and thus the automated calculation of phase behaviour without bias of prior knowledge. Predicting these is of interest in areas from high-energy materials to organic electronics and pharmaceuticals.
2. Demonstrate the application of Nested Sampling to atomistic models of flexible molecules of various sizes, from small molecules and short chained alkanes to oligomeric substances, to study their full phase diagram, including the melt and crystalline phases. Apart from gaining thermodynamic knowledge of the studied systems, there is also scope to inform the improvement or development of the employed models.
3. Disseminate the software tools and promote the technique via showcase publications, conference/workshop presentations and online examples/tutorials.

During the project, you will develop a strong foundation in state-of-the-art computational and statistical methods in materials modelling, gaining hands-on experience with a wide range of sampling techniques and Monte Carlo approaches, incorporating uncertainty quantification. You will work with different families of interatomic potential models for molecular systems - from coarse-grained representations to classical atomistic force fields, and potentially machine-learned models.

In addition, you will gain practical experience with high-performance computing environments, working with large data sets, and apply robust software engineering principles, and develop code optimised for parallel computing skills that are highly transferable to both academic and industrial research.

These skills position you for careers in AI research, computational materials science, national laboratories, tech industry or academic research. The HetSys training provides a foundation for these skills through dedicated courses and cohort activities.

We require at least a II(i) honours degree at BSc or an integrated masters degree (e.g. MPhys, MChem, MSci, MEng etc.) in a physical sciences, mathematics or engineering discipline. We do not accept applications from existing PhD holders.

If you are an overseas candidate please check here that you hold the equivalent grades before applying.

For postgraduate study in HetSys, the term “overseas” or “international” student refers to anyone who does not qualify for UK home fee status. This includes applicants from the European Union (EU), European Economic Area (EEA), and Switzerland, unless they hold settled or pre-settled status under the UK’s EU Settlement Scheme.