Wiley Interdisciplinary Reviews: Computational Molecular Science最新文献

The two-electron reduced density matrix (2RDM) carries enough information to evaluate the electronic energy of a many-electron system. The variational 2RDM (v2RDM) approach seeks to determine the 2RDM directly, without knowledge of the wave function, by minimizing this energy with respect to variations in the elements of the 2RDM, while also enforcing known N-representability conditions. In this tutorial review, we provide an overview of the theoretical underpinnings of the v2RDM approach and the N-representability constraints that are typically applied to the 2RDM. We also discuss the semidefinite programming (SDP) techniques used in v2RDM computations and provide enough Python code to develop a working v2RDM code that interfaces to the libSDP library of SDP solvers.

This article is categorized under:

The cover image is based on the Advanced Review Simplifi ed quantum chemistry methods to evaluate non-linear optical properties of large systems by Sarah Löffelsender et al., https://doi.org/10.1002/wcms.1695

Ultrafast electron dynamics have made rapid progress in the last few years. With Jellyfish, we now introduce a program suite that enables to perform the entire workflow of an electron-dynamics simulation. The modular program architecture offers a flexible combination of different propagators, Hamiltonians, basis sets, and more. Jellyfish can be operated by a graphical user interface, which makes it easy to get started for nonspecialist users and gives experienced users a clear overview of the entire functionality. The temporal evolution of a wave function can currently be executed in the time-dependent configuration interaction method (TDCI) formalism, however, a plugin system facilitates the expansion to other methods and tools without requiring in-depth knowledge of the program. Currently developed plugins allow to include results from conventional electronic structure calculations as well as the usage and extension of quantum-compute algorithms for electron dynamics. We present the capabilities of Jellyfish on three examples to showcase the simulation and analysis of light-driven correlated electron dynamics. The implemented visualization of various densities enables an efficient and detailed analysis for the long-standing quest of the electron–hole pair formation.

This article is categorized under:

Molecules are complex dynamic objects that can exist in different molecular forms (conformations, tautomers, stereoisomers, protonation states, etc.) and often it is not known which molecular form is responsible for observed physicochemical and biological properties of a given molecule. This raises the problem of the selection of the correct molecular form for machine learning modeling of target properties. The same problem is common to biological molecules (RNA, DNA, proteins)—long sequences where only key segments, which often cannot be located precisely, are involved in biological functions. Multi-instance machine learning (MIL) is an efficient approach for solving problems where objects under study cannot be uniquely represented by a single instance, but rather by a set of multiple alternative instances. Multi-instance learning was formalized in 1997 and motivated by the problem of conformation selection in drug activity prediction tasks. Since then MIL has found a lot of applications in various domains, such as information retrieval, computer vision, signal processing, bankruptcy prediction, and so on. In the given review we describe the MIL framework and its applications to the tasks associated with ambiguity in the representation of small and biological molecules in chemoinformatics and bioinformatics. We have collected examples that demonstrate the advantages of MIL over the traditional single-instance learning (SIL) approach. Special attention was paid to the ability of MIL models to identify key instances responsible for a modeling property.

This article is categorized under:

Peptide-based drugs offer high specificity, potency, and selectivity. However, their inherent flexibility and differences in conformational preferences between their free and bound states create unique challenges that have hindered progress in effective drug discovery pipelines. The emergence of AlphaFold (AF) and Artificial Intelligence (AI) presents new opportunities for enhancing peptide-based drug discovery. We explore recent advancements that facilitate a successful peptide drug discovery pipeline, considering peptides' attractive therapeutic properties and strategies to enhance their stability and bioavailability. AF enables efficient and accurate prediction of peptide-protein structures, addressing a critical requirement in computational drug discovery pipelines. In the post-AF era, we are witnessing rapid progress with the potential to revolutionize peptide-based drug discovery such as the ability to rank peptide binders or classify them as binders/non-binders and the ability to design novel peptide sequences. However, AI-based methods are struggling due to the lack of well-curated datasets, for example to accommodate modified amino acids or unconventional cyclization. Thus, physics-based methods, such as docking or molecular dynamics simulations, continue to hold a complementary role in peptide drug discovery pipelines. Moreover, MD-based tools offer valuable insights into binding mechanisms, as well as the thermodynamic and kinetic properties of complexes. As we navigate this evolving landscape, a synergistic integration of AI and physics-based methods holds the promise of reshaping the landscape of peptide-based drug discovery.

This article is categorized under:

This review presents the theoretical background concerning simplified quantum chemistry (sQC) methods to compute non-linear optical (NLO) properties and their applications to large systems. To evaluate any NLO responses such as hyperpolarizabilities or two-photon absorption (2PA), one should evidently perform first a ground state calculation and compute its response. Because of this, methods used to compute ground states of large systems are outlined, especially the xTB (extended tight-binding) scheme. An overview on approaches to compute excited state and response properties is given, emphasizing the simplified time-dependent density functional theory (sTD-DFT). The formalism of the eXact integral sTD-DFT (XsTD-DFT) method is also introduced. For the first hyperpolarizability, 2PA, excited state absorption, and second hyperpolarizability, a brief historical review is given on early-stage semi-empirical method applications to systems that were considered large at the time. Then, we showcase recent applications with sQC methods, especially the sTD-DFT scheme to large challenging systems such as fluorescent proteins or fluorescent organic nanoparticles as well as dynamic structural effects on flexible tryptophan-rich peptides and gramicidin A. Thanks to the sTD-DFT-xTB scheme, all-atom quantum chemistry methodologies are now possible for the computation of the first hyperpolarizability and 2PA of systems up to 5000 atoms. This review concludes by summing-up current and future method developments in the sQC framework as well as forthcoming applications on large systems.

This article is categorized under:

Nucleation is the initial step in the formation of crystalline materials from solutions. Various factors, such as environmental conditions, composition, and external fields, can influence its outcomes and rates. Indeed, controlling this rate-determining step toward phase separation is critical, as it can significantly impact the resulting material's structure and properties. Atomistic simulations can be exploited to gain insight into nucleation mechanisms—an aspect difficult to ascertain in experiments—and estimate nucleation rates. However, the microscopic nature of simulations can influence the phase behavior of nucleating solutions when compared to macroscale counterparts. An additional challenge arises from the inadequate timescales accessible to standard molecular simulations to simulate nucleation directly; this is due to the inherent rareness of nucleation events, which may be apparent in silico at even high supersaturations. In recent decades, molecular simulation methods have emerged to circumvent length- and timescale limitations. However, it is not always clear which simulation method is most suitable to study crystal nucleation from solution. This review surveys recent advances in this field, shedding light on typical nucleation mechanisms and the appropriateness of various simulation techniques for their study. Our goal is to provide a deeper understanding of the complexities associated with modeling crystal nucleation from solution and identify areas for further research. This review targets researchers across various scientific domains, including materials science, chemistry, physics and engineering, and aims to foster collaborative efforts to develop new strategies to understand and control nucleation.

This article is categorized under:

The resolution-of-the-identity (RI) or density fitting (DF) approximation for the electron repulsion integrals (ERIs) has become a standard component of accelerated and reduced-scaling implementations of first-principles Gaussian-type orbital electronic-structure methods. The Cholesky decomposition (CD) of the ERIs has also become increasingly deployed across quantum chemistry packages in the last decade, even though its early applications were mostly limited to high-accuracy methods such as coupled-cluster theory and multiconfigurational approaches. Starting with a summary of the basic theory underpinning both the CD and RI/DF approximations, thus underlining the extremely close relation of the CD and RI/DF techniques, we provide a brief and largely chronological review of the evolution of the CD approach from its birth in 1977 to its current state. In addition to being a purely numerical procedure for handling ERIs, thus providing robust and computationally efficient approximations to the exact ERIs that have been found increasingly useful on modern computer platforms, CD also offers highly accurate approaches for generating auxiliary basis sets for the RI/DF approximation on the fly due to the deep mathematical connection between the two approaches. In this review, we aim to provide a concise reference of the main techniques employed in various CD approaches in electronic structure theory, to exemplify the connection between the CD and RI/DF approaches, and to clarify the state of the art to guide new implementations of CD approaches across electronic structure programs.

This article is categorized under:

Retrosynthesis is the cornerstone of organic chemistry, providing chemists in material and drug manufacturing access to poorly available and brand-new molecules. Conventional rule-based or expert-based computer-aided synthesis has obvious limitations, such as high labor costs and limited search space. In recent years, dramatic breakthroughs driven by deep learning have revolutionized retrosynthesis. Here we aim to present a comprehensive review of recent advances in AI-based retrosynthesis. For single-step and multi-step retrosynthesis both, we first introduce their goal and provide a thorough taxonomy of existing methods. Afterwards, we analyze these methods in terms of their mechanism and performance, and introduce popular evaluation metrics for them, in which we also provide a detailed comparison among representative methods on several public datasets. In the next part, we introduce popular databases and established platforms for retrosynthesis. Finally, this review concludes with a discussion about promising research directions in this field.

This article is categorized under:

All processes involving molecular systems entail a balance between associated enthalpic and entropic changes. Molecular dynamics simulations of the end-points of a process provide in a straightforward way the enthalpy as an ensemble average. Obtaining absolute entropies is still an open problem and most commonly pathway methods are used to obtain free energy changes and thereafter entropy changes. The kth nearest neighbor (kNN) method has been first proposed as a general method for entropy estimation in the mathematical community 20 years ago. Later, it has been applied to compute conformational, positional–orientational, and hydration entropies of molecules. Programs to compute entropies from molecular ensembles, for example, from molecular dynamics (MD) trajectories, based on the kNN method, are currently available. The kNN method has distinct advantages over traditional methods, namely that it is possible to address high-dimensional spaces, impossible to treat without loss of resolution or drastic approximations with, for example, histogram-based methods. Application of the method requires understanding the features of: the kth nearest neighbor method for entropy estimation; the variables relevant to biomolecular and in general molecular processes; the metrics associated with such variables; the practical implementation of the method, including requirements and limitations intrinsic to the method; and the applications for conformational, position/orientation and solvation entropy. Coupling the method with general approximations for the multivariable entropy based on mutual information, it is possible to address high dimensional problems like those involving the conformation of proteins, nucleic acids, binding of molecules and hydration.

This article is categorized under: