ObjectiveThe widespread adoption of portable fundus cameras for primary care and community screening is hindered by limitations in current autofocus(AF) technologies. Image-based methods relying on sharpness evaluation require iterative searches, resulting in slow convergence, while projection-based techniques are susceptible to optical artifacts and calibration errors. To address these challenges, this study introduces a novel AF system based on direct wavefront sensing, designed to deliver simultaneous high speed, high precision, and operational robustness within the compact form factor essential for portable ophthalmic devices. MethodsOur approach fundamentally reimagines the AF process by directly measuring the ocular wavefront aberration. We developed a custom portable fundus camera integrating a miniaturized Shack-Hartmann wavefront sensor (SHWS) into the optical path. An 850 nm laser diode projects a point source onto the retina via oblique illumination to minimize corneal reflections. Light scattered from this spot carries the eye’s refractive error through the imaging optics and is directed to the SHWS, positioned at a plane optically conjugate to the primary color CMOS imaging sensor. A microlens array within the SHWS samples the incident wavefront, generating a pattern of focal spots on a CCD. Real-time centroid analysis of these spots provides a map of local wavefront slopes. These measurements are processed through a singular value decomposition (SVD) algorithm to fit a Zernike polynomial basis set, enabling real-time reconstruction of the wavefront phase. The defocus component (S) is extracted from the second-order Zernike coefficients, providing a direct, quantitative measure of the refractive error in diopters. This value serves as a precise error signal in a closed-loop control system, which commands a voice-coil actuated focusing lens to its null position in a single, deterministic step, eliminating the need for iterative search algorithms. ResultsComprehensive evaluation demonstrated the system’s high performance. Testing on a calibrated model eye (OEMI-7) established a highly linear relationship between the computed defocus S and the focusing lens position across a ±20 Diopter (D) compensation range, achievable within a 5 mm mechanical travel. The system achieved a focusing precision of 0.08 D, corresponding to an 18-fold improvement over a conventional projection spot-size method tested under identical conditions. The total focus acquisition time, encompassing wavefront measurement, computation, and lens actuation, averaged under 0.5 s. Clinical validation with 25 human volunteers (50 eyes, refractive range -15 D to +10 D) confirmed practical efficacy. The wavefront-sensing AF succeeded in 92% of attempts with a mean time of 0.5 s, substantially outperforming a projection-based benchmark which achieved only a 32% success rate with an average time of 4.25 s. The system provided instantaneous directional guidance and maintained stability during minor ocular movements. Objective assessment of image quality, via amplitude contrast of retinal vasculature, showed consistent and significant enhancement following AF correction across the entire tested diopter range. ConclusionThis work successfully implements and validates a direct wavefront-sensing autofocus paradigm for portable fundus cameras. By directly quantifying and compensating for the optical defocus aberration, this method bypasses the fundamental limitations of image-processing and projection-based techniques, enabling rapid, precise, and deterministic diopter compensation. The developed system delivers an exceptional combination of a wide operational range (±20 D), high accuracy (0.08 D), fast convergence (0.5 s), and a compact physical footprint. This technology provides a practical and high-performance focusing solution capable of enhancing the reliability, throughput, and diagnostic utility of portable retinal imaging in large-scale screening applications. Future efforts will be directed towards system cost optimization and performance adaptation for diverse ocular conditions.

ObjectiveProtein-protein interactions (PPIs) are fundamental to the execution of biological functions within living cells. However, traditional biochemical methods, such as co-immunoprecipitation (Co-IP), often fail to capture transient, weak, or membrane-associated interactions due to the stringent detergent requirements for cell lysis. Proximity labeling (PL) has emerged in recent years as a transformative technology for mapping the proteomes of specific subcellular compartments and identifying dynamic interactomes in situ. Golgi protein 73 (GP73, also known as GOLPH2), a resident type II Golgi transmembrane protein, is a well-recognized clinical biomarker for liver diseases, including hepatocellular carcinoma (HCC). Despite its clinical significance, the comprehensive physiological and pathological functions of GP73 remain partially understood. This study aims to establish an APEX2-mediated proximity labeling system specifically targeting GP73 to map its interactome in a living cellular environment, thereby providing new insights into its molecular roles and regulatory mechanisms. MethodsTo achieve spatial specificity, we first constructed a stable cell line expressing a fusion protein consisting of GP73 and the engineered soybean peroxidase APEX2. The localization of the GP73-APEX2 fusion protein was validated to ensure it correctly targeted the Golgi apparatus. The proximity labeling reaction was initiated by incubating the cells with biotin-phenol (BP) for 30 min, followed by a brief (1 min) treatment with1 mmol/L hydrogen peroxide (H₂O₂). This catalytic reaction converts BP into highly reactive, short-lived biotin-phenoxyl radicals that covalently attach to endogenous proteins within a small labeling radius of the GP73-APEX2 enzyme. Subsequently, the cells were quenched, and biotinylated proteins were enriched using high-affinity streptavidin-coated magnetic beads. The captured “neighbor” proteins were subjected to on-bead digestion and analyzed via liquid chromatography-tandem mass spectrometry (LC-MS/MS) for high-throughput identification. Rigorous bioinformatics analysis, including Gene Ontology (GO) enrichment, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, and protein-protein interaction network mapping, was performed to interpret the biological significance of the identified candidates. ResultsOur results demonstrate the successful establishment of a robust and sensitive APEX2-based proximity labeling system for GP73. We identified a total of 95 high-confidence interacting proteins that were significantly enriched in the GP73 proximity proteome compared to control groups. Bioinformatics analysis revealed that these interactors were predominantly associated with biological processes such as vesicular transport, protein localization, and, most notably, molecular functions related to “ribosome binding” and “translation regulation”. This suggested an unexpected role for the Golgi-resident GP73 in the cellular translation machinery. To validate these findings, we performed targeted biochemical assays which confirmed a direct interaction between GP73 and the subunits of the eukaryotic translation initiation factor 3 (eIF3) complex, specifically EIF3G and EIF3I. Furthermore, functional validation using the surface sensing of translation (SUnSET) assay—a non-radioactive method to monitor protein synthesis—revealed that the overexpression of GP73 significantly promoted global protein translation levels in the cell, whereas its depletion or inhibition resulted in reduced translation efficiency. ConclusionThis study successfully utilized APEX2-mediated proximity labeling to provide the first systematic map of GP73 interactome in living cells. Our findings uncover a novel, unconventional function of GP73 as a regulator of cellular protein translation, likely mediated through its interaction with the eIF3 complex. This discovery significantly broadens our understanding of the biological roles of GP73 beyond its traditional function in the Golgi apparatus and suggests that it may act as a bridge between Golgi-related trafficking and the protein synthesis machinery. Furthermore, the technical framework established in this study provides a valuable template for investigating other complex organelle-associated protein networks and resolving transient macromolecular interactions in various physiological and pathological contexts.

ObjectiveTo investigate the effects of the classical Chinese medicine compound prescription Erjingwan on the inflammatory response and apoptosis of skeletal muscle cells in a mouse model of sarcopenia and decipher the mechanism based on the silent information regulator 1 （SIRT1）/nuclear factor erythroid 2-related factor 2 （Nrf2）/heme oxygenase-1 （HO-1） signaling pathway. MethodsForty C57/BL6 male mice were randomized into a control group， a model group， and groups with different doses of Erjingwan （8，16，32 g·kg^-1）. The mouse model of sarcopenia was established by D-gal-induced skeletal muscle senescence. The body weight and grip strength of mice treated with different doses of Erjingwan were examined to evaluate their physiological functions. Hematoxylin-eosin （HE） staining and Masson staining were used to observe the pathological changes and fibrosis in the skeletal muscle of mice. Enzyme-linked immunosorbent assay （ELISA） was adopted to determine the levels of tumor necrosis factor-α （TNF-α） and interleukin-6 （IL-6） in the serum samples of mice， and biochemical tests were conducted to quantify the levels of superoxide dismutase （SOD）， malondialdehyde （MDA）， and glutathione （GSH） in the serum. The protein and mRNA levels of SIRT1， Nrf2， B-cell lymphoma （Bcl-2）， and Bcl-2-associated X protein （Bax） were determined by Western blot and Real-time fluorescence quantitative polymerase chain reaction （Real-time PCR）， respectively. ResultsAfter 4 weeks of drug intervention， the model group exhibited significant reductions in body weight and grip strength （P0.01） compared with the control group. Compared with the model group， all doses of Erjingwan increased the body weight in mice at week 8 （P0.01） and grip strength from week 6 （P0.01）. HE staining revealed clear muscle fiber structure in the control group， muscle fiber rupture and atrophy in the model group， and dose-dependent repair of muscle fiber structure in the Erjingwan groups. Masson staining showed minimal collagen fibers and mild fibrosis in the control group， collagen fiber proliferation and severe fibrosis in the model group， and collagen proliferation with dose-dependent inhibition of fibrosis in the Erjingwan groups. ELISA results showed that serum levels of TNF-α and IL-6 were elevated in the model group compared with those in the control group （P0.01）. After intervention， the low-dose Erjingwan group exhibited a decreased TNF-α level （P0.05）， while the medium and high-dose groups showed decreases in both TNF-α and IL-6 levels （P0.01）. Biochemical assays revealed that the model group had decreased SOD and GSH levels （P0.01） and an increased MDA level （P0.01） compared with the control group. The medium and high-dose Erjingwan groups exhibited increases in SOD and GSH levels （P0.01） and decreases in MDA level （P0.01）， compared with the model group. WB and Real-time PCR results showed that compared with the control group， the model group presented down-regulated protein and mRNA levels of SIRT1， Nrf2， HO-1， and Bcl-2 in the muscle tissue （P0.01） and up-regulated protein and mRNA levels of Bax （P0.01）. Compared with the model group， Erjingwan at different doses up-regulated the protein levels of SIRT1， Nrf2， HO-1， and Bcl-2 （P0.01） and down-regulated the protein and mRNA levels of Bax （P0.01） in the muscle tissue. Low-dose Erjingwan elevated the mRNA levels of Nrf2 and HO-1 （P0.05， P0.01）， and medium and high-dose Erjingwan up-regulated the mRNA levels of SIRT1， Nrf2， HO-1， and Bcl-2 （P0.01）. ConclusionErjingwan reduced the content of inflammatory factors in skeletal muscle cells， improved the antioxidant capacity， and attenuated pathological changes and fibrosis in the muscle of the mouse model of sarcopenia by regulating the SIRT1/Nrf2/HO-1 pathway， inflammatory response， and apoptosis network.

ObjectiveTo investigate the effects of the classical Chinese medicine compound prescription Erjingwan on the inflammatory response and apoptosis of skeletal muscle cells in a mouse model of sarcopenia and decipher the mechanism based on the silent information regulator 1 （SIRT1）/nuclear factor erythroid 2-related factor 2 （Nrf2）/heme oxygenase-1 （HO-1） signaling pathway. MethodsForty C57/BL6 male mice were randomized into a control group， a model group， and groups with different doses of Erjingwan （8，16，32 g·kg^-1）. The mouse model of sarcopenia was established by D-gal-induced skeletal muscle senescence. The body weight and grip strength of mice treated with different doses of Erjingwan were examined to evaluate their physiological functions. Hematoxylin-eosin （HE） staining and Masson staining were used to observe the pathological changes and fibrosis in the skeletal muscle of mice. Enzyme-linked immunosorbent assay （ELISA） was adopted to determine the levels of tumor necrosis factor-α （TNF-α） and interleukin-6 （IL-6） in the serum samples of mice， and biochemical tests were conducted to quantify the levels of superoxide dismutase （SOD）， malondialdehyde （MDA）， and glutathione （GSH） in the serum. The protein and mRNA levels of SIRT1， Nrf2， B-cell lymphoma （Bcl-2）， and Bcl-2-associated X protein （Bax） were determined by Western blot and Real-time fluorescence quantitative polymerase chain reaction （Real-time PCR）， respectively. ResultsAfter 4 weeks of drug intervention， the model group exhibited significant reductions in body weight and grip strength （P0.01） compared with the control group. Compared with the model group， all doses of Erjingwan increased the body weight in mice at week 8 （P0.01） and grip strength from week 6 （P0.01）. HE staining revealed clear muscle fiber structure in the control group， muscle fiber rupture and atrophy in the model group， and dose-dependent repair of muscle fiber structure in the Erjingwan groups. Masson staining showed minimal collagen fibers and mild fibrosis in the control group， collagen fiber proliferation and severe fibrosis in the model group， and collagen proliferation with dose-dependent inhibition of fibrosis in the Erjingwan groups. ELISA results showed that serum levels of TNF-α and IL-6 were elevated in the model group compared with those in the control group （P0.01）. After intervention， the low-dose Erjingwan group exhibited a decreased TNF-α level （P0.05）， while the medium and high-dose groups showed decreases in both TNF-α and IL-6 levels （P0.01）. Biochemical assays revealed that the model group had decreased SOD and GSH levels （P0.01） and an increased MDA level （P0.01） compared with the control group. The medium and high-dose Erjingwan groups exhibited increases in SOD and GSH levels （P0.01） and decreases in MDA level （P0.01）， compared with the model group. WB and Real-time PCR results showed that compared with the control group， the model group presented down-regulated protein and mRNA levels of SIRT1， Nrf2， HO-1， and Bcl-2 in the muscle tissue （P0.01） and up-regulated protein and mRNA levels of Bax （P0.01）. Compared with the model group， Erjingwan at different doses up-regulated the protein levels of SIRT1， Nrf2， HO-1， and Bcl-2 （P0.01） and down-regulated the protein and mRNA levels of Bax （P0.01） in the muscle tissue. Low-dose Erjingwan elevated the mRNA levels of Nrf2 and HO-1 （P0.05， P0.01）， and medium and high-dose Erjingwan up-regulated the mRNA levels of SIRT1， Nrf2， HO-1， and Bcl-2 （P0.01）. ConclusionErjingwan reduced the content of inflammatory factors in skeletal muscle cells， improved the antioxidant capacity， and attenuated pathological changes and fibrosis in the muscle of the mouse model of sarcopenia by regulating the SIRT1/Nrf2/HO-1 pathway， inflammatory response， and apoptosis network.

OBJECTIVE To provide a scientific basis for the quality evaluation and clinical application of Qingwen hufei granules. METHODS Fourteen batches of Qingwen hufei granules were used as samples to establish high-performance liquid chromatography （HPLC） fingerprints using the Similarity Evaluation System for Chromatographic Fingerprint of Traditional Chinese Medicine （2012 Edition）. The chromatographic peaks were identified and the similarity was evaluated. Cluster analysis （CA）， principal component analysis （PCA）， and orthogonal partial least squares-discriminant analysis （OPLS-DA） were used to conduct chemical pattern recognition analysis on the 14 batches of samples. Meanwhile， the contents of neochlorogenic acid （NGA）， chlorogenic acid （CHA）， cryptochlorogenic acid （CGA）， forsythoside A （FTA）， 3，5-O-dicaffeoylquinic acid （3，5-O- DA）， 4，5-O-dicaffeoylquinic acid （4，5-O-DA）， and angoroside C （AGC） in the samples were determined by HPLC. RESULTS The methodological investigation results of both the fingerprint and the content determination complied with the relevant requirements. Fourteen common peaks were indicated in the HPLC fingerprints of the 14 batches of samples， and 7 of them were identified [NGA （peak 2）， CHA （peak 3）， CGA （peak 5）， FTA （peak 11）， 3，5-O-DA （peak 12）， 4，5-O-DA （peak 13）， and AGC （peak 14）]； the similarity of each sample was greater than 0.94. The results of CA and PCA showed that the samples could be classified into 3 categories； the results of OPLS-DA indicated that peak 4 （unknown）， peak 11 （FTA）， peak 8 （unknown）， peak 9 （unknown）， and peak 1 （unknown） were the differential components. The content ranges of NGA， CHA， CGA， 3，5-O-DA， FTA， 4，5-O-DA and AGC in the 14 batches of samples were 0.210 4-0.458 7， 0.269 1-0.506 3， 0.228 1-0.461 1， 0.443 9-1.044 6， 0.066 7-0.155 7， 0.062 8-0.143 8， and 0.057 4-0.105 7 mg/g， respectively. CONCLUSIONS The HPLC fingerprint and multi-component content determination methods established in this study are efficient and reliable， and can be used for the quality evaluation of Qingwen hufei granules.

A quantitative method for the analysis of the multi-component contents in Marsdenia tenacissimae was established, and the quality differences were evaluated by principal component analysis (PCA), orthogonal partial least squares-discriminant analysis (OPLS-DA), factor analysis (FA) and weighted technique for order preference by similarity to ideal solution (TOPSIS) method. The contents of chlorogenic acid, cryptochlorogenic acid, sinapic acid, tenacigenoside A, tenacissoside G, tenacissoside I, tenacissoside H, drevogenin A, betulinic acid and lupeol were determined by HPLC wavelength switching method. At the same time, the contents of alcohol-soluble extract and total ash were detected. PCA, OPLS-DA and FA methods were used to identify the origin of M. tenacissimae from different producing areas. According to the OPLS-DA model, the index weight was determined to construct the weighted TOPSIS evaluation model. The qualities of M. tenacissimae from different producing areas were analyzed by model scoring results. The contents of 12 indexes in 18 batches of M. tenacissimae varied to different degrees, and the repeatability and accuracy of the test method were satisfactory. PCA analysis divided 18 batches of M. tenacissimae into three categories. OPLS-DA identified five main potential quality markers, including tenacissoside A, tenacissoside I, lupeol, tenacissoside H and chlorogenic acid. The evaluation results of FA and weighted TOPSIS method were consistent, which showed that the quality of M. tenacissimae from Yunnan and Guizhou was better. The established multi-component quantitative analysis method is accurate and reliable, the chemometrics model has strong predictive ability, and the evaluation results of FA and weighted TOPSIS method are scientific and objective. The combination of the four methods can clearly determine the qualities of M. tenacissimae from different producing areas.

BACKGROUND:High-frequency repetitive transcranial magnetic stimulation has garnered significant attention due to its potential non-invasive benefits in modulating brain function.However,no studies have comprehensively analyzed the current research landscape and development trends of this field from a macroscopic perspective.OBJECTIVE:To explore research hotspots,current trends,and emerging frontiers in the field of high-frequency repetitive transcranial magnetic stimulation through visualized analysis.METHODS:Data were collected from the Web of Science Core Collection database from January 1,2014 to November 15,2024.CiteSpace was used for analyzing publication volume,collaborations among countries/regions,institutions and authors,citation analysis of journals and co-cited literature,as well as disciplinary distribution.Additionally,keyword co-occurrence,clustering,and burst analyses were conducted,and visualized knowledge maps were generated.RESULTS AND CONCLUSION:A total of 860 articles were included.The publication volume of high-frequency repetitive transcranial magnetic stimulation showed an overall upward trend from 2014 to 2022,followed by a decline from 2022 to 2024.China had the highest publication volume,while Ghent University ranked as the most productive institution.Universities acted as the most high-output institutions.Chris Baeken from Ghent University was identified as the most prolific author.Collaboration among leading authors and institutions worldwide remained limited.The main research hotspots in this field were associated with keywords such as depression,stroke,neuropathic pain,and Parkinson's disease.Burst keywords focused on mild cognitive impairment,reflecting a diversification in research directions.The overall research activity in high-frequency repetitive transcranial magnetic stimulation continues to rise,with primary focuses on its clinical applications for psychiatric and neurological disorders,as well as explorations of its underlying mechanisms.Future research may focus on optimizing treatment parameters for targeting different brain regions in clinical applications and expanding its applications and mechanisms across various domains.

BACKGROUND:High-frequency repetitive transcranial magnetic stimulation has garnered significant attention due to its potential non-invasive benefits in modulating brain function.However,no studies have comprehensively analyzed the current research landscape and development trends of this field from a macroscopic perspective.OBJECTIVE:To explore research hotspots,current trends,and emerging frontiers in the field of high-frequency repetitive transcranial magnetic stimulation through visualized analysis.METHODS:Data were collected from the Web of Science Core Collection database from January 1,2014 to November 15,2024.CiteSpace was used for analyzing publication volume,collaborations among countries/regions,institutions and authors,citation analysis of journals and co-cited literature,as well as disciplinary distribution.Additionally,keyword co-occurrence,clustering,and burst analyses were conducted,and visualized knowledge maps were generated.RESULTS AND CONCLUSION:A total of 860 articles were included.The publication volume of high-frequency repetitive transcranial magnetic stimulation showed an overall upward trend from 2014 to 2022,followed by a decline from 2022 to 2024.China had the highest publication volume,while Ghent University ranked as the most productive institution.Universities acted as the most high-output institutions.Chris Baeken from Ghent University was identified as the most prolific author.Collaboration among leading authors and institutions worldwide remained limited.The main research hotspots in this field were associated with keywords such as depression,stroke,neuropathic pain,and Parkinson's disease.Burst keywords focused on mild cognitive impairment,reflecting a diversification in research directions.The overall research activity in high-frequency repetitive transcranial magnetic stimulation continues to rise,with primary focuses on its clinical applications for psychiatric and neurological disorders,as well as explorations of its underlying mechanisms.Future research may focus on optimizing treatment parameters for targeting different brain regions in clinical applications and expanding its applications and mechanisms across various domains.

[摘 要] 目的：探讨X连锁锌指蛋白（ZFX）通过Nectin-4表达影响食管鳞状细胞癌（ESCC）进展的分子机制及其对PI3K/AKT信号通路的激活作用。方法：收集2022年8月至2023年7月在南充市中心医院手术切除的30对ESCC组织及癌旁组织标本。采用人食管上皮细胞HET-1A及ESCC细胞KYSE-30、KYSE-150、KYSE-410、KYSE-510和TE-1。基于2018年至2019年收集的6例ESCC配对标本转录组测序数据筛选Nectin-4，通过TIMER2.0数据库、RT-qPCR和WB法、免疫组织化学（IHC）检测ESCC组织及细胞中Nectin-4的表达水平。采用shRNA技术在KYSE-410和KYSE-510细胞中敲低Nectin-4，通过CCK-8、克隆形成、划痕愈合及Transwell实验检测敲低Nectin-4对细胞增殖、迁移和侵袭能力的影响，采用WB法检测敲低Nectin-4对细胞PI3K/AKT通路及EMT相关蛋白表达水平的影响。通过生物信息学预测并结合双萤光素酶报告基因实验，鉴定并验证ZFX作为Nectin-4的上游转录调控因子。结果：综合分析显示，ESCC组织及细胞系中Nectin-4表达显著高于癌旁组织及HET-1A细胞（均P < 0.01）。功能研究表明，敲低Nectin-4显著抑制KYSE-410和KYSE-510细胞的增殖活性、克隆形成能力，以及迁移和侵袭能力（均P < 0.01）。机制方面，敲低Nectin-4时细胞中E-cadherin表达上调、N-cadherin下调（均P < 0.01），并抑制PI3K/AKT通路相关蛋白的磷酸化水平（P < 0.01）。结论： ZFX通过上调Nectin-4表达激活PI3K/AKT信号通路促进ESCC进展，为ESCC的治疗提供了潜在的新靶点。

In recent years, the application of artificial intelligence (AI) in the field of biology has witnessed remarkable advancements. Among these, the most notable achievements have emerged in the domain of protein structure prediction and design, with AlphaFold and related innovations earning the 2024 Nobel Prize in Chemistry. These breakthroughs have transformed our ability to understand protein folding and molecular interactions, marking a pivotal milestone in computational biology. Looking ahead, it is foreseeable that the accurate prediction of various physicochemical properties of proteins—beyond static structure—will become the next critical frontier in this rapidly evolving field. One of the most important protein properties is thermodynamic stability, which refers to a protein’s ability to maintain its native conformation under physiological or stress conditions. Accurate prediction of protein stability, especially upon single-point mutations, plays a vital role in numerous scientific and industrial domains. These include understanding the molecular basis of disease, rational drug design, development of therapeutic proteins, design of more robust industrial enzymes, and engineering of biosensors. Consequently, the ability to reliably forecast the stability changes caused by mutations has broad and transformative implications across biomedical and biotechnological applications. Historically, protein stability was assessed via experimental methods such as differential scanning calorimetry (DSC) and circular dichroism (CD), which, while precise, are time-consuming and resource-intensive. This prompted the development of computational approaches, including empirical energy functions and physics-based simulations. However, these traditional models often fall short in capturing the complex, high-dimensional nature of protein conformational landscapes and mutational effects. Recent advances in machine learning (ML) have significantly improved predictive performance in this area. Early ML models used handcrafted features derived from sequence and structure, whereas modern deep learning models leverage massive datasets and learn representations directly from data. Deep neural networks (DNNs), graph neural networks (GNNs), and attention-based architectures such as transformers have shown particular promise. GNNs, in particular, excel at modeling spatial and topological relationships in molecular structures, making them well-suited for protein modeling tasks. Furthermore, attention mechanisms enable models to dynamically weigh the contribution of specific residues or regions, capturing long-range interactions and allosteric effects. Nevertheless, several key challenges remain. These include the imbalance and scarcity of high-quality experimental datasets, particularly for rare or functionally significant mutations, which can lead to biased or overfitted models. Additionally, the inherently dynamic nature of proteins—their conformational flexibility and context-dependent behavior—is difficult to encode in static structural representations. Current models often rely on a single structure or average conformation, which may overlook important aspects of stability modulation. Efforts are ongoing to incorporate multi-conformational ensembles, molecular dynamics simulations, and physics-informed learning frameworks into predictive models. This paper presents a comprehensive review of the evolution of protein thermodynamic stability prediction techniques, with emphasis on the recent progress enabled by machine learning. It highlights representative datasets, modeling strategies, evaluation benchmarks, and the integration of structural and biochemical features. The aim is to provide researchers with a structured and up-to-date reference, guiding the development of more robust, generalizable, and interpretable models for predicting protein stability changes upon mutation. As the field moves forward, the synergy between data-driven AI methods and domain-specific biological knowledge will be key to unlocking deeper understanding and broader applications of protein engineering.