Wiley Interdisciplinary Reviews: Computational Molecular Science最新文献

Time-resolved photoelectron spectroscopy is a powerful pump-probe technique which can probe nonadiabatic dynamics in molecules. Interpretation of the experimental signals however requires input from theoretical simulations. Advances in electronic structure theory, nonadiabatic dynamics, and theory to calculate the ionization yields, have enabled accurate simulation of time-resolved photoelectron spectra leading to successful applications of the technique. We review the basic theory and steps involved in calculating time-resolved photoelectron spectra, and highlight successful applications.

This article is categorized under:

This paper focuses on the computationally assisted design of molecularly imprinted polymers (MIP), emphasizing the selected strategies and chosen methods of approach. In summary, this paper provides an overview of the MIP fabrication procedure, focusing on key factors and challenges, where the fabrication of MIP includes a step-by-step process with extensive experimental procedures. This brings challenges in optimizing experimental conditions, such as the selection of monomer, cross-linker, and their relevant molar ratios to the template and solvent. Next, the principles of computational methods are elucidated to explore their potential applicability in solving the challenges. The computational approach can tackle the problems and optimize the MIP's design. Finally, the atomistic, quantum mechanical (QM), and combined methods in the recent research studies are overviewed with stress on strategies, analyses, and results. It is demonstrated that optimization of pre-polymerization mixture by employing simulations significantly reduces the trial-and-error experiments. Besides, higher selectivity and sensitivity of MIP are observed. The polymerization and resulting binding sites by computational methods are considered. Several models of binding sites are formed and analyzed to assess the affinities representing the sensitivity and selectivity of modeled cavities. Combined QM/atomistic methods showed more flexibility and versatility for realistic modeling with higher accuracy. This methodological advancement aligns with the principles of green chemistry, offering cost-effective and time-efficient solutions in MIP design.

This article is categorized under:

Molecular dynamics (MD) simulations can provide detailed insights into complex molecular systems, such as DNA, at high resolution in space and time. Using current computer architectures, time scales of tens of microseconds are feasible with contemporary all-atom force fields. However, these timescales are insufficient to accurately characterize large conformational transitions in DNA and compare calculations to experimental data. This review discusses the advantages and drawbacks of two simulation approaches to overcome the timescale challenge. The first approach is based on adding biasing potentials to the system to drive transitions. Umbrella sampling, steered MD, and metadynamics are examples of these methods. A key challenge of such methods is the necessity of selecting one or a few efficient coordinates, commonly referred to as collective variables (CVs), along which to apply the biasing potential. The path-metadynamics methodology addresses this issue by finding the optimal route(s) between states in a multi-dimensional CV space. The second strategy is path sampling, which focuses MD simulations on the transitions. The assumption is that even though transitions between states are rare, they are generally fast. Stopping the simulations as soon as they reach a stable state can significantly increase simulation efficiency. We introduce these methods on the two-dimensional Müller–Brown potential. DNA applications are featured for two different processes: the Watson–Crick–Franklin to Hoogsteen transition in adenine–thymine base pairs and the binding of a DNA-binding protein domain to DNA.

This article is categorized under:

Generative AI is rapidly transforming the frontier of research in computational structural biology. Indeed, recent successes have substantially advanced protein design and drug discovery. One of the key methodologies underlying these advances is diffusion models (DM). Diffusion models originated in computer vision, rapidly taking over image generation and offering superior quality and performance. These models were subsequently extended and modified for uses in other areas including computational structural biology. DMs are well equipped to model high dimensional, geometric data while exploiting key strengths of deep learning. In structural biology, for example, they have achieved state-of-the-art results on protein 3D structure generation and small molecule docking. This review covers the basics of diffusion models, associated modeling choices regarding molecular representations, generation capabilities, prevailing heuristics, as well as key limitations and forthcoming refinements. We also provide best practices around evaluation procedures to help establish rigorous benchmarking and evaluation. The review is intended to provide a fresh view into the state-of-the-art as well as highlight its potentials and current challenges of recent generative techniques in computational structural biology.

This article is categorized under:

Stereochemical effects significantly influence chemical processes, yet it is not well understood if they are a leading source of uncertainty in combustion modeling. Stereochemistry influences a combustion model (i) at the earliest stage of its construction when mapping the reaction network, (ii) in the computation of individual thermochemical and rate parameters, and (iii) in the prediction of combustion observables. The present work reviews the importance of enumerating stereochemical species and reactions at each of these steps. Further, it analyzes the separate influence of several types of stereochemistry, including geometric, optical, and fleeting transition state diastereomers. Three reaction networks serve to examine which stages of low-temperature oxidation are most affected by stereochemistry, including the first and second oxidation of n-butane, the third oxidation of n-pentane, and the early stages of pyrolysis of 1- and 2-pentene. The 149 reactions in the n-butane mechanism are expanded to 183 reactions when accounting for diastereomerism. Each of these 183 reactions is parameterized with ab initio kinetics computations to determine that, for the n-butane mechanism, the median factor of diastereomeric deviation is 3.5 at 360 K for rate constants and as high as 1.6 for mechanism reactivity, in terms of ignition delay times, as opposed to a mechanism without stereochemical expansion.

This article is categorized under:

Ab-initio quantum chemistry simulations are essential for understanding electronic structure of molecules and materials in almost all areas of chemistry. A broad variety of electronic structure theories and implementations has been developed in the past decades to hopefully solve the many-body Schrödinger equation in an approximate manner on modern computers. In this review, we present recent progress in advancing low-rank electronic structure methodologies that rely on the wavefunction sparsity and compressibility to select the important subset of electronic configurations for both weakly and strongly correlated molecules. Representative chemistry applications that require the many-body treatment beyond traditional density functional approximations are discussed. The low-rank electronic structure theories have further prompted us to highlight compressive and expressive principles that are useful to catalyze idea of quantum learning models. The intersection of the low-rank correlated feature design and the modern deep neural network learning provides new feasibilities to predict chemically accurate correlation energies of unknown molecules that are not represented in the training dataset. The results by others and us are discussed to reveal that the electronic feature sets from an extremely low-rank correlation representation, which is very poor for explicit energy computation, are however sufficiently expressive for capturing and transferring electron correlation patterns across distinct molecular compositions, bond types and geometries.

This article is categorized under:

The field of liquid-phase and solid-state high-pressure chemistry has exploded since the advent of the diamond anvil cell, an experimental technique that allows the application of pressures up to several hundred gigapascals. To complement high-pressure experiments, a large number of computational tools have been developed. These techniques enable the simulation of chemical systems, their sizes ranging from single atoms to infinitely large crystals, under high pressure, and the calculation of the resulting structural, electronic, and spectroscopic changes. At the most fundamental level, computational methods using carefully tailored wall potentials allow the analytical calculation of energies and electronic properties of compressed atoms. Molecules and molecular clusters can be compressed either via mechanochemical approaches or via more sophisticated computational protocols using implicit or explicit solvation approaches, typically in combination with density functional theory, thus allowing the simulation of pressure-induced chemical reactions. Crystals and other periodic systems can be routinely simulated under pressure as well, both statically and dynamically, to predict the changes of crystallographic data under pressure and high-pressure crystal structure transitions. In this review, the theoretical foundations of the available computational tools for simulating high-pressure chemistry are introduced and example applications demonstrating the strengths and weaknesses of each approach are discussed.

This article is categorized under:

Accurate and rapid prediction of protein–ligand interactions (PLIs) is the fundamental challenge of drug discovery. Deep learning methods have been harnessed for this purpose, yet the insufficient generalizability of PLI prediction prevents their broader impact on practical applications. Here, we highlight the significance of PLI model generalizability by conceiving PLIs as a function defined on infinitely diverse protein–ligand pairs and binding poses. To delve into the generalization challenges within PLI predictions, we comprehensively explore the evaluation strategies to assess the generalizability fairly. Moreover, we categorize structure-based PLI models with leveraged strategies for learning generalizable features from structure-based PLI data. Finally, we conclude the review by emphasizing the need for accurate pose-predicting methods, which is a prerequisite for more accurate PLI predictions.

This article is categorized under:

Pedersen TB, Lehtola S, Fdez. Galván I, Lindh R. The versatility of the Cholesky decomposition in electronic structure theory. WIREs Comput Mol Sci. 2024; 14(1):e1692. https://doi.org/10.1002/wcms.1692.

We apologize for this error and thank Prof. L. De Vico for bringing this to our attention.

Amyloid proteins are characterized by their tendency to aggregate into amyloid fibrils, which are often associated with devastating diseases. Aggregation pathways typically involve unfolding or misfolding of monomeric proteins and formation of transient oligomers and protofibrils before the final aggregation product is formed. The conformational dynamics and polymorphic and volatile nature of these aggregation intermediates make their characterization by experimental techniques alone insufficient and also require computational approaches. Over the past 25 years, the size of simulated amyloid aggregation systems and the length of these simulations have increased significantly. These advances are discussed here. The review includes simulation approaches that model the aggregating peptides or proteins at both the all-atom and coarse-grained levels, use molecular dynamics simulations or Monte Carlo sampling to simulate the conformational changes, and present results for various amyloid peptides and proteins ranging from Lys-Phe-Phe-Glu (KFFE) as the smallest system to $��\mathrm{A\beta }�$ as an intermediate-sized peptide to α-synuclein. The presentation of the history of amyloid aggregation simulations concludes with a discussion of where the future of these simulations may lie.

This article is categorized under: