Article
Physics, Multidisciplinary
Raul Borsche, Axel Klar, Mattia Zanella
Summary: This research explores the application of a hierarchical description of traffic flow control by driver-assist vehicles, including lane changing dynamics. Lane-dependent feedback control strategies are implemented at the vehicle level through Boltzmann-type equations, and a system of first order macroscopic equations describing the evolution of densities along the lanes is determined through a suitable closure strategy. Numerical examples are presented to illustrate the features of the proposed hierarchical approach.
PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS
(2022)
Article
Chemistry, Multidisciplinary
Dean J. Tantillo
Summary: Differences in entropies can guide kinetic selectivity, but understanding and modeling these differences at the molecular level is challenging due to multiple vibrational states, dynamically accessible pathways, and contributions from different conformations/configurations.
JOURNAL OF THE AMERICAN CHEMICAL SOCIETY
(2022)
Article
Astronomy & Astrophysics
Chris Hamilton
Summary: The unshielded nature of gravity leads to inherent inhomogeneity in stellar systems, requiring the use of angle-action variables in kinetic theory. Collective interactions and polarization effects can enhance or suppress the relaxation of star clusters and galaxies. A recent angle-action generalization of the Balescu-Lenard equation accounts for both inhomogeneity and collective effects, providing a simpler derivation using Rostoker's superposition principle, which connects the BL picture to classical two-body relaxation theory.
MONTHLY NOTICES OF THE ROYAL ASTRONOMICAL SOCIETY
(2021)
Article
Computer Science, Interdisciplinary Applications
Michael A. Calicchia, Ehsan Atefi, John C. Leylegian
Summary: This study updates a method for generating small, accurate kinetic models for computational fluid dynamics programs, balancing model fidelity with time constraints through optimizing rate constant parameters, resulting in more accurate models obtained in less time.
ENGINEERING WITH COMPUTERS
(2022)
Article
Chemistry, Medicinal
Marcel Rak, Roberta Tesch, Lena M. Berger, Ekaterina Shevchenko, Monika Raab, Amelie Tjaden, Rezart Zhubi, Dimitrios-Ilias Balourdas, Andreas C. Joerger, Antti Poso, Andreas Kra, Lewis Elson, Aleksandar Luc, Thales Kronenberger, Thomas Hanke, Klaus Strebhardt, Mourad Sanhaji, Stefan Knapp
Summary: Salt-inducible kinases 1-3 (SIK1-3) are important regulators of cellular homeostasis. This study presents a structure-based approach to improve the selectivity of inhibitors targeting SIK kinases, resulting in the development of a valuable tool compound, MR22, which showed excellent selectivity and phenotypic effects in ovarian cancer cells.
EUROPEAN JOURNAL OF MEDICINAL CHEMISTRY
(2023)
Article
Physics, Fluids & Plasmas
Petra Papez, Tomaz Urbic
Summary: In this study, two-dimensional models of simple alcohols were designed using the Mercedes-Benz model of water. Monte Carlo simulations were conducted to investigate the structural and thermodynamic properties of these models, and it was found that they exhibited similar trends as observed in experiments. This research provides a simple testing ground to study the competition between polar and non-polar effects within molecules and their physical properties.
Article
Chemistry, Physical
Shashikant Kumar, Babak Sadigh, Siya Zhu, Phanish Suryanarayana, Sebastian Hamel, Brian Gallagher, Vasily Bulatov, John Klepeis, Amit Samanta
Summary: In this paper, a method for training kinetic energy functional models is proposed, using physically relevant terms from the literature and linear or nonlinear regression methods to obtain fitting coefficients. The predictive capabilities of these models are assessed using various model systems, and it is shown that high accuracy models can be generated using data from exact-exchange KSDFT calculations.
JOURNAL OF CHEMICAL PHYSICS
(2022)
Article
Engineering, Civil
Michael L. Follum, Jacob D. Scott, James W. Lewis, Joseph L. Gutenson, Ahmad A. Tavakoly, Mark D. Wahl
Summary: We propose a method to create a continuous topobathymetric dataset for large-scale applications using simple hydraulic models, gridded digital elevation model (DEM) data, a stream network dataset, and average flow rate. The method burns-in one of three bathymetric profiles into auto-generated cross-sections from the DEM and uses spatial averaging to create a realistic initial elevation surface. The accuracy of the method was tested in the Mississippi River Basin and showed promising results.
JOURNAL OF HYDROLOGY
(2023)
Article
Engineering, Mechanical
G. Tsialiamanis, N. Dervilis, D. J. Wagg, K. Worden
Summary: Machine learning has influenced the modelling of various phenomena, including structural dynamics, but problem-specific algorithms often struggle with data scarcity. To overcome this, a combination of physics-based and machine learning approaches has been developed. This study aims to promote the use of models that learn relationships from similar populations of phenomena, inspired by transferable physics-based models. Two meta-learning algorithms, model-agnostic meta-learning (MAML) and conditional neural processes (CNP), are utilized to achieve data-driven models. These algorithms outperform traditional machine learning algorithms in approximating quantities of interest and exhibit behavior similar to neural networks and Gaussian processes.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2023)
Article
Chemistry, Physical
Anupam Anand Ojha, Saumya Thakur, Surl-Hee Ahn, Rommie E. Amaro
Summary: Recent advances in computational power and algorithms have extended the time scales of molecular dynamics (MD) simulations. However, MD simulations still face limitations in observing conformational transitions associated with biomolecular processes. To address this challenge, enhanced sampling techniques such as the weighted ensemble (WE) method have been developed to estimate kinetic rate constants by sampling transitions between metastable states using weighted trajectories. In this study, deep-learned kinetic modeling approaches are introduced to extract statistically relevant information from short MD trajectories and provide a well-sampled initial state distribution for WE simulations. This hybrid approach overcomes statistical bias and produces a refined free energy landscape closer to the steady state, enabling efficient sampling of kinetic properties.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2023)
Article
Chemistry, Physical
Pallab Dutta, Neelanjana Sengupta
Summary: Computer simulations are increasingly used to analyze the thermo-kinetic information of protein kinase structural transformation. However, high computational cost and lack of standard protocols for low dimensional physical descriptors encoding transition-important system features pose challenges. In this study, we construct physically meaningful, orthogonal collective variables to preserve the slow modes of the system, based on the distinct catalytic activities of Abelson tyrosine kinase. Estimation of global partition function along appropriate physical descriptors using modified Expectation Maximized Molecular Dynamics method reveals excellent agreement with experimentally known rate-limiting dynamics and activation energy. Further development and applications are discussed.
Article
Multidisciplinary Sciences
Martin-I. Trappe, Ryan A. Chisholm
Summary: Ecology currently lacks a holistic approach to model phenomena across temporal and spatial scales, but density functional theory provides a promising computational framework to address this challenge.
NATURE COMMUNICATIONS
(2023)
Article
Chemistry, Physical
Peter G. G. Bolhuis, Z. Faidon Brotzakis, Bettina G. G. Keller
Summary: Empirical force fields in molecular dynamics simulations are often optimized to reproduce structural and thermodynamic properties, but rarely consider the rates of interconversion between metastable states. This study introduces a framework that combines dynamical observables and molecular model parameters to optimize force field parameters based on the constraint of matching predicted and experimental rate constants. The approach combines statistical mechanics of trajectories with path reweighting methods and automatically selects the solutions that perturb the entire path ensemble the least, following the maximum entropy principle. The methodology is demonstrated on simple test systems and provides physical insights into the sensitivity of the model to kinetics and has broad implications.
JOURNAL OF CHEMICAL PHYSICS
(2023)
Article
Biology
Mattia Zanella
Summary: Understanding the impact of collective social phenomena on epidemic dynamics is crucial for effectively containing the disease spread. This study proposes a mathematical model that assesses the interplay between opinion polarization and disease evolution. The model shows that the spread of the disease is closely related to consensus dynamics distribution and the emergence of opinion polarization. Numerical investigations confirm the model's ability to describe various phenomena related to the spread of an epidemic.
BULLETIN OF MATHEMATICAL BIOLOGY
(2023)
Article
Thermodynamics
Bin Yang
Summary: The accurate and robust kinetic models are crucial in combustion chemistry research. Reliable experimental data and advanced diagnostic techniques play essential roles in model development. New strategies like ASSM and ANN-HDMR can reduce model dimensionality and computational cost significantly, while global-sensitivity based experimental design methods can guide kinetics-information-enriched data generation. The computational framework OptEx provides a new means for integrating experimental data with mechanism development, design, and optimization to develop reliable kinetic models more efficiently and effectively.
PROCEEDINGS OF THE COMBUSTION INSTITUTE
(2021)
Article
Immunology
Jamie B. Spangler, Eleonora Trotta, Jakub Tomala, Ariana Peck, Tracy A. Young, Christina S. Savvides, Stephanie Silveria, Petra Votavova, Joshua Salafsky, Vijay S. Pande, Marek Kovar, Jeffrey A. Bluestone, K. Christopher Garcia
JOURNAL OF IMMUNOLOGY
(2018)
Article
Multidisciplinary Sciences
Trung Hai Nguyen, Arien S. Rustenburg, Stefan G. Krimmer, Hexi Zhang, John D. Clark, Paul A. Novick, Kim Branson, Vijay S. Pande, John D. Chodera, David D. L. Minh
Article
Biochemical Research Methods
Keri A. McKiernan, Anna K. Koster, Merritt Maduke, Vijay S. Pande
PLOS COMPUTATIONAL BIOLOGY
(2020)