Nucleation is a fundamental step in the formation of new materials, with nucleation rates governing phase-transition pathways and outcomes that influence material properties across chemistry, physics, and biology. Nevertheless, extracting nucleation rates remains challenging: experiments are limited by the microscopic length scales involved, while molecular simulations are hindered by the rare-event nature of nucleation and the large system sizes required to sample low-supersaturation regimes typical of experiments and industrial processing. Here, we build on the established thermodynamic correspondence between stable clusters in small, closed systems and critical clusters in open systems and develop a multicomponent extension, the Critical Cluster Equivalence Principle, that transforms this correspondence into a practical workflow for calculating nucleation rates. We employ this equivalence to determine solvent-mediated NaCl crystal nucleation driving forces and rates over a wide range of supersaturations (SNaCl ∈ [1.5, 4]) and further validate the framework by benchmarking against curvature-corrected homogeneous nucleation rates for argon vapor condensation. This has been achieved using a small number of brute-force, finite-size simulations in which cluster-size statistics and monomer-exchange dynamics in the steady-state map directly onto critical clusters in macroscopic systems. By combining this multicomponent correspondence with the computational approach developed here, the current investigation has obtained nucleation rates in excellent agreement with experiments and enhanced sampling simulations. This approach provides a unified and generalizable tool for predicting nucleation rates with minimal computational effort, enabling routine application across a wide range of material systems.
Lun-Na Li, Fabienne Bachtiger, A. Finney et al.· Journal of the American Chem...· 1 citation
Recent years have witnessed considerable progress in computer-aided drug discovery, driven by the incorporation of computational technologies within both academic and pharmaceutical environments. This evolution is marked by a significant accumulation of data pertaining to detailed three-dimensional structural information, ligand properties, and their interactions with therapeutic targets. The augmentation of computational capabilities and the accessibility of extensive chemical libraries containing billions of drug-like small molecules have further facilitated this transition. To effectively utilize these resources, it is imperative to employ rapid computing methods for virtual screening, which encompass structure-driven in silico screening across vast molecular spaces, supported by efficient recurrent profiling techniques. Furthermore, advancements in deep learning methodologies are required to improve the accuracy of prediction concerning target functionalities and ligand characteristics, even when complete receptor structures are not available. Here, this review discusses the expansion of chemical space, advanced virtual screening, deep learning, molecular dynamics (MD) simulations, Absorption, Distribution, Metabolism, Excretion, and Toxicity (ADMET) modelling, and challenges for drug discovery. It examines how experimental validation integrates computational predictions with laboratory testing for effective candidate selection. Finally, it outlines future research directions for artificial intelligence-driven, ultra-high performance computing (UHPC) in drug discovery, offering new prospects for the economical creation of safer and more efficacious molecule-level therapies.
Lunna Li, Lianna D. Soriano, W. M. Kedir et al.· Molecular Biomedicine· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
