Session 10

Software development and application for material design based on the statistical mechanics theory of liquids

Norio Yoshida
on  Thu, 14:50 for  20min
N. Yoshida
1. Graduate School of Informatics, Nagoya University
noriwo@nagoya-u.ac.jp

Abstract

Biological environments are composed of complex mixtures of water, electrolytes, proteins, lipids, and various other biomolecules. To design biocompatible materials that function effectively in such environments, it is essential to account for the intricate physicochemical properties of these multi-component solutions. Traditional simulation approaches often face challenges when trying to capture solvent effects at molecular resolution, especially in heterogeneous and biologically relevant conditions.

To address this, we have developed software based on the three-dimensional reference interaction site model (3D-RISM), a statistical mechanics theory of molecular solvation, referred as the reference interaction site model integrated calculator (RISMiCal).[1-4] The 3D-RISM theory enables the efficient and accurate prediction of solvation structures and thermodynamic properties by solving integral equations for solvent distribution functions around solutes. Our implementation focuses on biomolecular systems, allowing for the analysis of solvation and interactions of proteins, polymers, and other macromolecules in solution.

This software provides a powerful platform for understanding the structure-function relationship of biomolecules in complex solvents and has applications in material design, drug discovery, and biotechnology. In this presentation, we will introduce the theoretical background and capabilities of the software, highlighting recent advancements in methodology, including improvements in numerical solvers, hybrid approaches with molecular dynamics, and enhanced sampling techniques.

Furthermore, we will discuss recent developments that integrate 3D-RISM theory with machine learning methods. By combining physics-based solvation theory with data-driven models, we aim to accelerate the prediction and screening of material properties, opening new avenues in the design of smart, adaptive, and biocompatible materials.

References

[1] N. Yoshida, IOP Conf. Series: Materials Science and Engineerging, 773 (2020),012062 (DOI: 10.1088/1757-899X/773/1/012062)

[2] N. Yoshida, J. Chem. Info. Model., 57, (2017) 2646-2656 (DOI:10.1021/acs.jcim.7b00389)

[3] Y. Maruyama, N. Yoshida, J. Comput. Chem., (2024) 45, 1470-1482 (DOI: 10.1002/jcc.27340)

[4] K. Kanemaru, T. Yamaguchi, N. Sakumichi, T. Sakai, N. Yoshida J. Phys. Chem. B. (2025) 129, 9769-9780 (DOI: 10.1021/acs.jpcb.5c03519 )