Equilibrium and non-equilibrium thermodynamics in drug repurposing: Machine learning-guided discovery of high-affinity WEE1 kinase inhibitors.

Single-ion hydration free energies stringently test molecular models for aqueous ions, but quantitative comparison is complicated by the interplay of short-range ion-water interactions, long-range electrostatics, many-body polarization, and nuclear quantum effects (NQE). Here, we compute hydration free energies for alkali-metal cations (Li+-Cs+) and halide anions (F--I-) using MB-nrg ion potentials in MB-pol water. Free energies are evaluated with a staged alchemical cycle in which ion-water interactions are introduced sequentially (charge, polarization, and explicit 2-body and 3-body terms), enabling stable sampling of each contribution. To perform robust charging transformations within the MBX electrostatics framework, we implement a soft-core Coulomb scaling and evaluate free-energy changes using finite-difference thermodynamic integration. Model fidelity is assessed primarily by comparison to experimental within-series reference differences, ΔΔGhyd, which reduce sensitivity to single-ion reference conventions. Across both ion families, MB-nrg/MB-pol reproduces the expected monotonic weakening of hydration with increasing ionic size, yielding RMSE values of 2.12 and 3.39 kcal/mol for the alkali-metal and halide ΔΔGhyd series, respectively. Including NQE produces only small shifts in ΔGhyd but lowers these RMSE values to 1.61 and 2.31 kcal/mol, respectively, which are consistent with very small NQE-induced changes in ion-water structure. Ion-water radial distribution functions show that electrostatics rapidly establishes the first hydration shell, while the explicit many-body corrections relax short-range overstructuring toward the fully interacting reference distribution and NQE slightly broadens and reduces first-shell structuring. Overall, these results establish a practical foundation for predictive hydration free-energy calculations with data-driven many-body potentials in bulk water, and motivate extensions to interfacial hydration and nanoconfined aqueous environments.

In-cylinder pressure is a key parameter for evaluating combustion processes and engine performance in spark-ignition engines. However, acquiring high-resolution pressure data over a wide range of operating conditions, particularly under varying spark advance (SA), is costly and technically challenging, which limits its practical application. To address this issue, this study proposes two artificial neural network (ANN)-based methods for in-cylinder pressure reconstruction using data from a three-cylinder gasoline engine under different spark advance conditions. Both methods employ crank angle and spark advance as input features. The first method (ANN-P) directly predicts the in-cylinder pressure profile, achieving a coefficient of determination (R2) exceeding 0.99 on both training and validation datasets, with a root mean square error (RMSE) below 0.13 bar. The model accurately reproduces the pressure evolution throughout the compression, combustion, and expansion processes and enables reliable estimation of indicated mean effective pressure (IMEP). The second method (ANN-HRR) adopts an indirect strategy by first predicting the heat release rate (HRR) and subsequently reconstructing the pressure trace through thermodynamic integration based on a single-zone model. This approach avoids error amplification associated with numerical differentiation and demonstrates improved accuracy in predicting combustion phasing metrics, such as CA10 and CA50. The results indicate that both methods effectively capture the influence of spark timing on combustion characteristics and peak pressure. While ANN-P provides higher accuracy in pressure reconstruction, ANN-HRR offers superior performance in characterizing combustion features. Overall, this study presents a cost-effective and accurate framework for combustion diagnostics, performance calibration, and control optimization of gasoline engines.

Molecular photoswitches enable spatiotemporal photocontrol of protein function, but their design requires high target selectivity and large light-dependent changes in binding affinity and/or efficacy. These properties are especially difficult to optimize in membrane receptors due to membrane-protein interactions. Computational design remains challenging because few benchmarks rigorously compare free-energy methods against experiment. Here, we establish such a benchmark for photoswitchable antagonists of β-adrenergic receptors, exemplifying most successful designs in the photopharmacology of class A G protein-coupled receptors (GPCRs) to date. We evaluated widely used free-energy methods for predicting how substituents and chirality affect light-responsive binding and subtype selectivity. Thermodynamic integration shows the best agreement with experiment, followed by umbrella sampling, whereas metadynamics and end-point methods perform poorly. Our simulations reveal interactions stabilizing cis OP2 in β2-AR and the key role of PHE289 in isomer-specific binding, consistent with mutagenesis data. Overall, this work provides a robust computational framework for GPCR photopharmacology.

In the present work, we have utilized “alchemical” approaches to study the thermodynamic properties of four alkane molecules: pentane, hexane, heptane, and octane in water. We have used thermodynamic integration (TI) and free energy perturbation (FEP) based methods: TI, TI‐cubic, Bennett acceptance ration (BAR), and Multistate Bennett acceptance ratio (MBAR)to estimate the solvation free energy of alkane molecules in water at 300 K for 21 distinct coupling parameter values (λ). For each alkane molecule, the estimated values of solvation free energy using the different methods are in close agreement, which ensures the reliability of our study. Convergence of the calculation has also been examined through time series plots in both forward and reverse directions. Our study shows that the solvation free energy increases with increasing size of the alkane molecules, which is also supported by estimation of solvent accessible surface area (SASA). The self diffusion coefficients of both solutes and solvent molecules; and shear viscosity of the systems have been estimated at 293 and 300 K temperatures. The self diffusion coefficient of the alkane molecules decreases with increase in size and shear viscosity increases with chain length as expected. The estimated values of diffusion and shear viscosity are in close agreement with the previously reported values.

Towards greener shipping: Thermodynamic modelling and feasibility of methanol reforming, CO₂ capture and port-side CO2 utilisation for UK maritime corridors

Hydrogen-fueled Solid Oxide Fuel Cell - Gas Turbine (SOFC-GT) hybrid engines for aircraft propulsion: Comprehensive thermodynamic heat integration study

Assessing thermodynamic performance and system integration of solar-powered batch reverse osmosis for seawater desalination

Free energy calculations have evolved from specialized theoretical tools to essential components of modern molecular design, particularly in pharmaceutical research where alchemical binding free energy methods now guide lead optimization decisions. Despite significant methodological advances over the past two decades, including statistically optimal estimators, enhanced sampling techniques, multi-dimensional nonequilibrium protocols, and improved force fields, current methods achieve reliable accuracy only for favorable systems involving rigid proteins and uncharged ligands. Challenges include inadequate sampling of protein flexibility, systematic biases from classical force fields for charged species, and convergence failures for systems with slow conformational transitions, limiting broader application to challenging but industrially relevant targets. This review provides a comprehensive analysis of free energy calculation methodologies, from foundational approaches (thermodynamic integration, Bennett acceptance ratio) to advanced enhanced sampling strategies (replica exchange, metadynamics) and cutting-edge developments including machine learning integration. Current accuracy benchmarks are assessed through systematic analysis of community challenges and industrial validation studies, identifying specific failure modes and their underlying physical origins. This review emphasizes practical method selection criteria and realistic accuracy expectations while also examining how machine learning approaches-including neural network potentials, automated collective variable discovery, and active learning protocols-address limitations in sampling efficiency and force field accuracy.

The temperature-induced orthorhombic to cubic phase transition in hbox {Li}_2hbox {C}_2 is a prototypical example of a solid to solid phase transformation between an ordered phase, which is well described within the phonon theory, and a dynamically disordered phase with rotating molecules, for which the standard phonon theory is not applicable. The transformation in hbox {Li}_2hbox {C}_2 happens from a phase with directionally ordered hbox {C}_2 dimers to a structure, where they are dynamically disordered. We provide a description of this transition by employing ab initio molecular dynamics (AIMD) based stress-strain thermodynamic integration on a deformation path that connects the ordered and dynamically disordered phases. The free energy difference between the two phases is obtained. The entropy that stabilizes the dynamically disordered cubic phase is captured by the behavior of the stress on the deformation path.

In this study, we combined field observations and microtextural analyses of the altered rocks with mass balance calculations, thermodynamic modelling, and stable isotopes to establish the formation and equilibria environments of alteration minerals and hydrothermal fluids at Mt. Ruapehu (New Zealand). Results indicate that the secondary minerals follow different precipitation sequences as a function of temperature and water type (i.e., either acid at the proximal distant areas from the summit/crater, or distal areas affected with rain-glacial composition), interacting with the magmas of intermediate composition. Variations in composition and abundances of the hydrothermal paragenesis in Mt. Ruapehu determine the evolution of its acid–sulfate alteration that also envisage a crystallization-driven degassing. We conclude that the combination of the proposed petrologic-geochemical-isotopic approach, along with geological, geophysical and the regional and local tectonic features, can effectively be used for the assessment of volcanic-hydrothermal eruptions, interpretation of monitoring data (degassing) and/or flanks instabilities. • Integration of petrology, thermodynamics and stable isotopes of each equilibria environment of alteration minerals and hydrothermal fluids at Ruapehu (New Zealand). • Newly-grown alteration minerals result from water-rock reactions within different hydrothermal environments of the volcanic edifice. • Precise physico-chemical conditions of the connection between the plumbing system and the volcano's surface.

Type I collagen is the main structural protein of vertebrates and forms molecular trimers from the COL1A1 and COL1A2 gene products, proα1(I) and proα2(I), during biosynthesis. Calcium ions are required for trimers to form. The amino acid sequence of the C-propeptide of collagen, which is removed before collagen fibril formation, initially drives heterotrimerization. Abnormal homotrimeric type I collagen is associated with age-related diseases including cancer, fibrosis, and musculoskeletal and cardiovascular conditions, but the circumstances under which the homotrimer may form are poorly understood. Here, we used molecular dynamics simulations of the C-propeptide protein structure to show that inter- and intrachain hydrogen bonding is affected by loss of calcium and that this leads chains to become destabilized, particularly at the interfaces of each chain. Loss of calcium resulted in increased distances between the cysteine residues that form interchain disulfide bonds, preventing the formation of these bonds. Pulling simulations and modeling of calcium dissociation from monomers showed that calcium ions were more strongly bound to the α1(I) than the α2(I) chain. However, enhanced sampling methods implied the α2(I) chain has a higher trimer affinity than a third α1(I) chain in the presence of structural calcium. To quantify assembly thermodynamics, we computed relative binding free energies by alchemical thermodynamic integration, demonstrating that α2(I)-specific residues at the interchain interface conferred a measurable thermodynamic advantage to trimer formation in the presence of calcium. Hence, although heterotrimerization is normally favored, in reduced calcium conditions the homotrimer can form by sequestering available calcium to the α1(I) chains. This study provides a molecular explanation for a calcium-based mechanism driving heterotrimerization versus homotrimerization of type I collagen.

The human metapneumovirus (HMPV) Fusion (F) glycoprotein is a high-priority target for "fusion-locking" agents that stabilize its metastable prefusion state. While monomeric catechins like EGCG are known antivirals, the molecular basis for the superior activity of structurally complex dimeric catechins remains poorly understood. We employed an advanced biophysical workflow, integrating 100 ns all-atom molecular dynamics (MD), free energy landscape (FEL) analysis, and MM/GBSA thermodynamic integration to decode the Structure-Dynamics Relationship (SDR) of 210 Camellia sinensis (Green tea) phytochemicals. The results reveal a "Galloylation-Driven Anchoring" mechanism: the galloyl moiety of prodelphinidin A2 3'-gallate provides critical electrostatic complementarity to the Asp325-Asp336 acidic ridge. FEL analysis quantitatively demonstrates that this anchoring leads to pronounced stabilization of the F protein in a deep, kinetically favored global minimum (ΔG = 9.357 kJ/mol), effectively raising the energy barrier for the fusogenic conformational shift. This study provides a comparative and mechanistically informed computational proof-of-concept for the use of dimeric natural scaffolds as precision fusion-locking agents, offering a roadmap for experimental biophysical validation. In this workflow, molecular docking was employed exclusively as a qualitative structure-based filtering step, while all quantitative conclusions regarding stabilization and binding energetics were derived from post-docking MD, FEL, and MM/GBSA analyses.

Protonation states of class-C $$\varvec{\beta}$$-lactamases determined by a combination of thermodynamic integration and unified free energy dynamics

We develop a method to fit high-temperature Gibbs free energy data for the development of interatomic potentials for atomic systems. The approach is based on Hamiltonian thermodynamic integration, enabling the identification of suitable potential parameters such that the system's free energy matches a specified target. The method can be readily combined with conventional fitting techniques for properties such as elastic tensors and liquid pair distribution functions. We validate the effectiveness of the approach using the Uhlenbeck-Ford model and embedded-atom method potentials for pure Ni phases and binary Fe1-xOx liquids under high-pressure and high-temperature conditions. Our framework provides an efficient strategy for incorporating free energy into interatomic potential fitting.

Standard harmonic-to-anharmonic thermodynamic integration (TI) is known to develop a near singularity in the integrand for solids exhibiting diffusive degrees of freedom, such as rotating functional groups or migrating defects. This pathology results in numerical challenges for estimating absolute free energies within a single thermodynamic cycle. In this work, we introduce a simple regularization that removes this singularity and yields a well-behaved integrand that can be accurately evaluated on a uniform grid. The approach-termed Regularized End point Gradient (REG) TI-is demonstrated on a model system and on predicting the relative stability of paracetamol polymorphs for which quasi-free methyl rotations lead to a near singularity in standard TI. We expect REG TI to simplify anharmonic free energy calculations for solids and to potentially enable their automation.

Amid the global energy structure's profound transformation and full implementation of the dual carbon strategy, optimal planning of integrated energy systems (IESs) is a cutting-edge energy topic. Most relevant studies rely on the first law of thermodynamics, focusing on energy conservation, while integration of energy quality, thermodynamics, and economics remains to be improved. This study takes the planning phase of IESs as the entry point and proposes an analytical method with exergy theory as the core and exergoeconomics as the tool. It proposes a closed-loop full-process planning framework of "theoretical modeling-scheme construction-economic analysis-scheme improvement" to develop a mature, replicable IES planning model, providing theoretical and technical support for the energy system's low-carbon transformation. Standard case simulations validate the efficacy of the closed-loop planning framework in enhancing system exergy efficiency and optimizing economic costs, providing engineering decision support and insights for multidisciplinary paradigm advancement.

Molecular dynamics study of M-Trifluoromethyl diphenyl diselenide binding to the μ-opioid receptor: A computational perspective on morphine-induced tolerance.

Context . Markov chain Monte Carlo (MCMC) excels at sampling complex posteriors, but traditionally lags behind nested sampling in accurate evidence estimation, which is crucial for model comparison in astrophysical problems. Aims . We introduce reddemcee , an adaptive parallel tempering ensemble sampler, aiming to close this gap by simultaneously presenting next-generation automated temperature-ladder adaptation techniques and robust, low-bias evidence estimators. Methods . reddemcee couples an affine-invariant stretch move with five interchangeable ladder-adaptation objectives – a uniform swap-acceptance rate, swap mean distance, Gaussian area overlap, a small Gaussian gap, and equalised thermodynamic length – implemented through a common differential update rule. Three evidence estimators are provided: curvature-aware thermodynamic integration (TI+), geometric-bridge stepping stones (SS+), and a novel hybrid algorithm that blends both approaches (H+). The performance and accuracy of the sampler are benchmarked on n-dimensional Gaussian shells, Gaussian egg-box, Rosenbrock functions, and the real exoplanet radial-velocity time-series dataset of HD 20794. Results . Across shells up to 15 dimensions, reddemcee achieves roughly 7 times the effective sampling speed of the best dynamic nested sampling configuration. The TI+, SS+, and H+ estimators recover estimates to within |∆ ln 𝒵|≲3% and supply realistic error bars with as few as six temperatures. In the HD 20794 case study, reddemcee reproduces literature model rankings and yields tighter yet consistent planetary parameters compared with dynesty , with evidence errors that track run-to-run dispersion. Conclusions . By unifying fast ladder adaptation with reliable evidence estimators, reddemcee delivers strong throughput and accurate evidence estimates, often matching, and occasionally surpassing, dynamic nested sampling, while preserving the rich posterior information that makes MCMC indispensable for modern Bayesian inference.

Background/Objectives: Detecting copy number variations (CNVs) from next-generation sequencing (NGS) is challenging, particularly in targeted sequencing panels, especially for cell-free DNA (cfDNA), where the signal is weak and noise is high. Methods: We present BayesCNV, a Bayesian hierarchical model for gene-level copy ratio estimation from targeted amplicon read depths compared to a CNV-neutral reference sample. The model provides posterior uncertainty for each gene and supports interpretable calling based on effect size and posterior confidence. The model also provides a principled quality-control strategy based on the marginal log likelihood of each sample, with low values indicating low confidence in the calls. BayesCNV uses thermodynamic integration, a technique to reliably estimate this quantity. We benchmark our method against two publicly available CNV callers using Seracare® reference samples with known CNVs on the OncoReveal® Core Lbx panel. Results: Our method achieves a sensitivity of 0.87 and specificity of 0.996, dramatically outperforming two competitor methods, IonCopy and DeviCNV. In a separate FFPE dataset using the OncoReveal® Essential Lbx panel, we show that the marginal log likelihood cleanly separates, degraded from high-quality samples, even when conventional sequencing QC metrics do not. Conclusions: BayesCNV provides accurate and interpretable gene-level CNV estimates and uncertainty quantification, along with an evidence-based quality control metric that improves robustness in targeted cfDNA workflows.