PhD Computational Statistical Physics, Imperial College London (funded by Marie Curie scholarship; advisor: Prof. Serafim Kalliadasis); MSc Theoretical Physics, Saint-Petersburg State University.


Peter is a cross-disciplinary researcher with a track record of publications in statistical physics, physics of fluids, applied mathematics and machine learning.

He is presently an Honorary Senior Research Associate at Imperial College, with effect from February 2024.

Peter built and supports a computational toolbox for multiscale modelling of statistical mechanical systems, fluids and interfaces. The toolbox implements integral-differential equations of equilibrium and dynamic density functional theory (DFT) in 1D/2D/3D using spectral collocation methods and finite element methods. The code is presently undergoing refactoring and will be soon publicly available. For his development of the DFT computational toolbox, in 2017 Peter was awarded a prestigious Sir William Wakeham Award of Imperial College London.

In short, DFT is a formulation of statistical mechanics in terms of operators acting on phase-space probability densities. By employing free-energy formalism, DFT offers an appealing computational device. It allows one to bridge scales between microscopic descriptions of dynamical systems, based on Hamiltonian mechanics and Liouville’s theorem, and macroscopic continuum-mechanical treatments, such as Navier-Stokes equations.

Within the Group, Peter works on machine learning approaches to inverse problems, numerical methods for integral-differential DFT equations, phase transitions and criticality of fluids at sub-capillary scales, physics of interfaces and non-equilibrium statistical mechanics.

Peter actively contributes to teaching and MSc/PhD student supervision at Imperial College. He is a Fellow of the UK Higher Education Academy (FHEA).

More details and a full list of publications can be found on Peter’s personal website

Recent highlights:

  1. Peter Yatsyshin, S. Kalliadasis, A. B. Duncan (2022) Physics-constrained Bayesian inference of state functions in classical density-functional theory Journal of Chemical Physics, 156 (7), 074105, DOI: 10.1063/5.0071629.
    [NeurIPS workshop]
  2. Peter Yatsyshin, S. Kalliadasis (2021) Surface nanodrops and nanobubbles: a classical density functional theory study
    Journal of Fluid Mechanics, 913 (1), DOI: 10.1017/jfm.2020.1167.
  3. Peter Yatsyshin, N. G. Fytas, E. Theodorakis (2020) Mixing-demixing transition in polymer-grafted spherical nanoparticles
    Soft Matter, 15, 703, DOI: 10.1039/c9sm01639b.


Fully microscopic DFT computation of nano-drops/bubbles (colour) against a properly calibrated macroscopic model of thin films (black lines). More details and insights on the connections between molecular-scale interactions and contact line hysteresis are in this JFM article