Article
Physics, Multidisciplinary
Alfred C. K. Farris, David P. Landau
Summary: The study investigated long sequences of the HP model of protein folding using replica exchange Wang-Landau sampling, revealing lower ground state energies compared to earlier simulations using PERM and replica exchange Monte Carlo, and extracting specific heat curves for the first time.
PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS
(2021)
Article
Computer Science, Interdisciplinary Applications
Stefan Schnabel, Wolfhard Janke
Summary: This paper demonstrates how the well-known Wang-Landau method can be modified to produce non-flat distributions, leading to increased efficiency for certain systems. Examples of such enhancements are provided.
COMPUTER PHYSICS COMMUNICATIONS
(2021)
Article
Chemistry, Physical
Joel A. K. L. Picard, Thomas Speck
Summary: Conventional gas-liquid phase transitions exhibit a coexistence line with a monotonic and positive slope, indicating that cooling leads to condensation. However, we investigate the opposite phenomenon of condensation of adsorbed organic molecules into dense domains upon heating. Using a simple lattice model, we study this process through Monte Carlo simulations, mean-field theory, and the analytical solution of the Ising model in two dimensions. Our findings are applicable to molecules with distinct conformations and different entropies or heat capacities.
JOURNAL OF CHEMICAL PHYSICS
(2023)
Article
Computer Science, Interdisciplinary Applications
Wanyok Atisattapong, Pasin Marupanthorn
Summary: This study investigated three alternative ensembles for estimating network reliability, concluding that random walks on structure function failed in highly reliable networks, while other methods performed efficiently for certain network types. The 1/t algorithm using random walks on probability space was the only ensemble that yielded accurate estimates for a dodecahedron network, but still failed at the highest level of network reliability. Further investigation is needed for methods that can reduce variance for large, highly reliable networks.
COMPUTER PHYSICS COMMUNICATIONS
(2021)
Article
Chemistry, Multidisciplinary
Yusuke Nanba, Michihisa Koyama
Summary: The unique features of alloy nanoparticles arise from the arrangement of elements within them, with different properties exhibited by solid solution and segregated configurations even with the same overall composition. The configuration space of alloy nanoparticles expands exponentially with an increase in constituent elements, making it difficult to estimate the configurational entropy. The method developed involving Wang-Landau sampling, density functional theory calculations, and multiple regression analysis is effective for probing the stable configurations of multinary alloy nanoparticles at a finite temperature.
BULLETIN OF THE CHEMICAL SOCIETY OF JAPAN
(2021)
Article
Chemistry, Multidisciplinary
Andrew Stannard, Marc Mora, Amy E. M. Beedle, Marta Castro-Lopez, Stephanie Board, Sergi Garcia-Manyes
Summary: Molecular fluctuations reveal the energy landscape of proteins, with variance analysis showing that unfolding and refolding transitions in proteins under mechanical forces result in changes in protein stiffness. The study demonstrates that the change in protein compliance with force-induced thermodynamically stable states is proportional to the protein's contour length increment, in line with the freely jointed chain model in polymer physics. These findings provide insights into the conformational dynamics of proteins under mechanical force which are crucial for mechanosensing and mechanotransduction.
Article
Polymer Science
Zhixing Huang, Yashasvi Bajaj, Jan-Michael Y. Carrillo, Yohei Nakanishi, Kiminori Uchida, Kazuki Mita, Takeshi Yamada, Tsukasa Miyazaki, Bobby G. Sumpter, Maya Endoh, Tadanori Koga
Summary: The interface between the polymer matrix and carbon fiber (CF) in carbon fiber reinforced polymers (CFRPs) significantly affects the macroscopic properties of CFRPs. This study characterized the bound polymer layer (BPL) on the CF surface using experimental and simulation techniques, and found that the crystallinity of the BPL was lower than the bulk, and the bound chains exhibited faster dynamics and formed a high-density region near the CF surface.
Article
Computer Science, Interdisciplinary Applications
Felipe Moreno, Sergio Davis, Joaquin Peralta
Summary: In this work, we developed an implementation of the Wang-Landau algorithm to find the density of states (DOS) in a given system. The implementation, using Python and C++ languages, takes advantage of libraries for energy computation and considers code parallelization for faster computation. The effectiveness and accuracy of the implementation were demonstrated through studies on various systems, providing a powerful and flexible tool for studying realistic matter models.
COMPUTER PHYSICS COMMUNICATIONS
(2022)
Article
Physics, Fluids & Plasmas
Meng Yao, Da Wang, Qiang-Hua Wang
Summary: In this paper, a method using the Wang-Landau algorithm in determinant quantum Monte Carlo to achieve flat-histogram sampling in configuration weight space is employed, significantly reducing the autocorrelation time in the Holstein model.
Article
Biology
Kei Moritsugu
Summary: In this study, a multiscale enhanced sampling (MSES) method based on variational autoencoder (VAE) is proposed, which effectively achieves enhanced sampling of protein structures by dynamically modeling in a reduced-dimensional subspace. By utilizing MD trajectories and VAE model, the structural features of both closed and open forms of the ribose-binding protein (RBP) were successfully transformed into interpolated data in the latent space, driving structural sampling at an atomistic resolution. The free energy surfaces on the latent space demonstrated the refinement from single basin to multiple closed and open basins, highlighting the utility of MD simulation and molecular mechanics force field in recovering accurate structural ensembles.
Article
Chemistry, Physical
Meredith M. Rickard, Haolin Luo, Ashley De Lio, Martin Gruebele, Taras V. Pogorelov
Summary: The cytoplasm affects the conformation of ATP, with ATP molecules bound to proteins in cells forming specific pitched conformations. These interactions may play functional roles when ATP interacts with protein surfaces.
JOURNAL OF PHYSICAL CHEMISTRY LETTERS
(2022)
Article
Chemistry, Medicinal
Kiyoto Aramis Tanemura, Susanta Das, Kenneth M. Merz
Summary: The study introduces an automated conformational clustering algorithm that reduces predefined cluster numbers or thresholds, while preserving geometric/energetic correlations. Automating conformational clustering may alleviate human biases and allow flexibility.
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2021)
Article
Computer Science, Artificial Intelligence
Oscar Mendez-Lucio, Mazen Ahmad, Ehecatl Antonio del Rio-Chanona, Jorg Kurt Wegner
Summary: DeepDock, developed by Mendez-Lucio and colleagues, utilizes geometric deep learning to predict binding conformations of ligands to protein targets, providing guidance for molecular optimization. The method learns a statistical potential based on distance likelihood, tailored for each ligand-target pair, and can reproduce experimental binding conformations effectively. This approach showcases the use of artificial intelligence to enhance structure-based drug design.
NATURE MACHINE INTELLIGENCE
(2021)
Article
Biochemical Research Methods
Namir Oues, Sarath Chandra Dantu, Riktaben Jigarkumar Patel, Alessandro Pandini
Summary: MDSubSampler is a Python library and toolkit for a posteriori subsampling of data from multiple trajectories, providing access to various sampling methods such as uniform, random, stratified, weighted, and bootstrapping sampling. It allows sampling under the constraint of preserving the original distribution of relevant geometrical properties, and can be used for simulations post-processing, noise reduction, and structures selection for ensemble docking.
Article
Computer Science, Artificial Intelligence
Prashant K. Gupta, Javier Andreu-Perez
Summary: The Wang-Mendel Approach (WMA) aims to improve the explainability of inference models by combining numerical and linguistic information. To address the limited capability of modeling linguistic information, we propose a novel Enhanced WMA that uses type-2 (T2) FSs for linguistic information modeling. We demonstrate the performance of Enhanced WMA through real-world applications and compare it with other methods.
JOURNAL OF EXPERIMENTAL & THEORETICAL ARTIFICIAL INTELLIGENCE
(2022)