Wiley Interdisciplinary Reviews: Computational Molecular Science最新文献

Nucleosomes serve as fundamental units of chromatin packaging and play a crucial role as central hubs in epigenetic regulation. Their positions throughout the genome are not random and follow certain patterns, influenced by DNA sequence, histone-DNA interactions, chromatin physical barriers, nucleosome sliding and unwrapping, and chromatin modifications. There are many experimental techniques for identifying nucleosome positions, but these methods often involve a trade-off between achieving high resolution and covering the entire genome. In this regard, computational approaches may offer a fast alternative, with the benefit of aiding experimental analysis by denoising data, refining nucleosome boundaries, and identifying features critical for nucleosome positioning. Moreover, computational predictions enable the integration of nucleosome positioning data with other genomic and epigenomic datasets, providing a more comprehensive view of chromatin organization and gene regulation. In this review, we focus on various nucleosome positioning methods, including experimental techniques of nucleosome boundaries identification and in silico methods of nucleosome positioning data denoising and prediction of nucleosome positioning from the DNA sequence.

Vibrational spectroscopy is commonly applied for investigating the chemical and physical properties of water and aqueous solutions. Ab initio spectroscopy methods are used to analyze experimental spectra, offering valuable insights into structural and dynamic properties. In cases where experimental data is limited or contentious for aqueous systems subjected to high pressure–temperature conditions or extreme spatial confinement, ab initio methods can provide guidance for experiments. Recent progress in algorithms and computational power has driven substantial development in ab initio spectroscopy. In this review, we summarize first principles methods for calculating dipole moments and electronic polarizabilities, as well as demonstrate the use of time correlation functions for calculating infrared (IR) and Raman spectra. Additionally, we summarize recent advances in machine learning methods developed to expedite spectrum calculations and discuss the existing challenges that require further advancements in the field.

Version 6.0 of the ORCA quantum chemistry program suite was released in July 2024. ORCA 6.0 is a major turning point in the history of the program since it represents a near complete rewrite of the code base that leads to: (1) major performance improvements, (2) a clean and highly efficient code base that greatly facilitates future development, (3) a large amount of new functionality, and (4) new interface capabilities that facilitate inter-operability with other quantum chemistry program packages. The article describes the most salient features of the program.

Coarse-graining (CG) is transforming the study of molecular systems, allowing researchers to explore by computer simulations larger and more complex structures than ever before. Continued advancements in CG techniques are making simulations more efficient, establishing this approach as a cornerstone for designing innovative materials and eco-friendly alternatives to traditional plastics. Additionally, CG methods are becoming indispensable for unraveling the complexities and functional mechanisms of large-scale macromolecular machines within cells. Yet, crafting an effective coarse-grained model demands a nuanced understanding of its advantages and limitations. Faster simulations come at the cost of molecular detail and accuracy in some properties, so that it is essential to balance computational efficiency with the specific needs of the system one wants to simulate. By asking the right questions, researchers can select models that offer the desired benefits while managing trade-offs. This article delves into the potential of different CG models and the compromises inherent in their adoption, highlighting their role in shaping the future of material science and biophysics.

Polymeric materials are intricate systems with unique properties across different length and time scales, presenting challenges in understanding the hierarchical features that govern their behavior. Advancing innovative polymeric systems requires a deep comprehension of these complexities. Dissipative particle dynamics (DPD), a mesoscale simulation technique, has proven instrumental in elucidating polymer behavior. Unlike molecular dynamics, which tracks individual molecules, DPD employs a coarse-graining approach, to describe molecular systems as particles interacting via soft potentials. Thanks to its computational efficiency, DPD has enabled researchers to numerically study several complex fluid applications in detail. Moreover, with the ever-increasing high-performance computing resources, it has become possible to tackle larger molecular systems beyond the nanoscale, typically micrometer-sized systems. An in-depth analysis of the theoretical foundations of DPD is presented, focusing on its methodology, mathematical formulations, and computational implementation. This review then explores various applications of DPD simulations for polymeric systems, demonstrating DPD's ability to accurately capture phenomena such as polymer self-assembly, polymer behavior in solutions and blends, charged polymers, polymer interfaces, polymer rheology, polymeric membranes, polymerization reactions, and polymeric composites. Overall, this review examines the adoption of DPD as a predictive modeling tool for polymeric materials, focusing on its key features and its integration with methods such as atomistic molecular dynamics to determine the interaction parameters. Building on these advancements, future directions for DPD include its potential applications in other systems like biological membranes, macromolecules, and shape-memory materials.

Transferring the success of the coupled-cluster expansion for single-determinant references to multireference cases remains a challenge. The main dilemma is a proper merge of the exponential ansatz, required for extensivity of the correlation energy, with a linear ansatz, required for an unbiased treatment of near-degenerate state interactions. We argue that the state interaction aspect is important and that therefore the Bloch equations are the necessary starting point for all true multireference coupled-cluster theories. Considering the aspect of spin-adaptation and orbital invariance, we arrive at internally contracted expansions, which indeed have a number of appealing formal properties, but also incur a tremendous increase in the complexity of the resulting working equations. The most striking property of internally contracted expansions is probably that a simple transformation of the reference space turns the multistate equations into state-specific equations without introducing further approximations. We discuss the present shortcomings and perspectives of the internally contracted multireference coupled-cluster theory and discuss issues like the completeness of the equations, alternative expansions using normal ordering, and perspectives for large active spaces and large molecules.