AIM:To compare the efficacy and safety of early combination therapy with ranibizumab and dexamethasone intravitreal implants versus ranibizumab monotherapy for the treatment of macular edema secondary to retinal vein occlusion(RVO-ME).METHODS: A retrospective cohort study was conducted on a total of 62 cases(64 eyes)of patients who were first diagnosed with RVO-ME at the Eye Centre of the Affiliated Hospital of Shandong Second Medical University between February 2022 and February 2023. The subjects were divided into two groups according to the different treatment regimens: 32 cases(34 eyes)in the monotherapy group received only ranibizumab [3+pro re nata(PRN)regimen], and 30 cases(30 eyes)in the combination therapy group were injected with ranibizumab once first, followed by dexamethasone intravitreal implant 3 wk later(1+DEX regimen). The best corrected visual acuity(BCVA), central retina thickness(CRT), foveal avascular zone(FAZ)area, macular vascular density(MVD)at the level of the deep vascular complex(DVC)of the retina, the incidence of ocular adverse effects, the number of drug injections, and the total cost between the two groups were compared before and after treatment.RESULTS: At 3 wk, 3 and 6 mo, and at the final follow-up of the two groups of patients, the improvement in BCVA, CRT, and MVD in the DVC layer was significantly better than that before treatment(all P<0.05); there were differences in the comparisons of BCVA and CRT between the two groups at 6 mo and the final follow-up(all P<0.05), and the increase in the number of letters of BCVA was the most pronounced in the combination therapy group at 6 mo of treatment. Statistical significant difference was observed in the comparison of MVD in the DVC layer between the two groups at 3 and 6 mo after treatment and at the final follow-up(all P<0.05). However, no significant change in FAZ area was evident before and after treatment in both groups(P>0.05). The combination therapy group exhibited a reduced number of injections and total cost in comparison to the monotherapy group. The combination therapy group exhibited a slightly higher incidence of high intraocular pressure and cataract progression compared to the monotherapy group, with no statistical significant difference(all P>0.05). Furthermore, no serious adverse events were observed in either group following treatment.CONCLUSION:Compared with ranibizumab alone, ranibizumab combined with dexamethasone intravitreal implant significantly improved vision, reduced macular edema, and lowered the frequency of injections and total treatment cost in patients with RVO-ME. CRT and MVD in the DVC layer are reliable prognostic indicators for patients with RVO-ME.

Objective: To analyze the change trends in the body shape indicators and proportions of students in Harbin from 2010 to 2024, and to investigate ergonomic compatibility of functional dimensions of school desks and chairs with current student shape indicators, so as to provide a reference for revising furniture standards of desks and chairs. Methods: Between September and November of both 2010 and 2024, a combination of convenience sampling and stratified cluster random sampling was conducted across three districts in Harbin, yielding samples of 6 590 and 6 252 students, respectively. Anthropometric shape indicators cluding height, sitting height, crus length, and thigh length-and their proportional changes were compared over the 15-year period. The 2024 data were compared with current standard functional dimensions of school furniture. The statistical analysis incorporated t-test and Mann-Whitney U- test. Results: From 2010 to 2024, average height increased by 1.8 cm for boys and 1.5 cm for girls; sitting height increased by 1.5 cm for both genders; crus length increased by 0.3 cm for boys and 0.4 cm for girls; and thigh length increased by 0.5 cm for both genders. The ratios of sitting height to height, and sitting height to leg length increased by less than 0.1 . The difference between desk chair height and 1/3 sitting height ranged from 0.4-0.8 cm. Among students matched with size 0 desks and chairs, 22.0% had a desk to chair height difference less than 0, indicating that the desk to chair height difference might be insufficient for taller students. The differences between seat height and fibular height ranged from -1.4 to 1.1 cm; and the differences between seat depth and buttock popliteal length ranged from -9.8 to 3.4 cm. Among obese students, the differences between seat width and 1/2 hip circumference ranged from -20.5 to -8.7 cm, while it ranged from -12.2 to -3.8 cm among non obese students. Conclusion Current furniture standards basically satisfy hygienic requirements; however, in the case of exceptionally tall and obese students, ergonomic accommodations such as adaptive seating allocation or personalized adjustments are recommended to meet hygienic requirements.

ObjectiveTo preliminarily explore the active components and target pathways of Angelicae Sinensis Radix-Polygonati Rhizoma （ASR-PR） herb pair in the treatment of simple obesity through network pharmacology and molecular docking， and to verify and investigate its mechanism of action via animal experiments. MethodsThe chemical constituents and targets of ASR and PR were predicted using the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform （TCMSP）. Targets related to simple obesity were identified by retrieving the GeneCards， Online Mendelian Inheritance in Man （OMIM）， Pharmacogenomics Knowledgebase （PharmGKB）， and DisGeNET databases. The intersection of drug and disease targets was used to construct an active component-target network using Cytoscape software. This network was imported into the STRING database to construct a protein-protein interaction （PPI） network， and topological analysis was conducted to identify core genes. Gene Ontology （GO） analysis and Kyoto Encyclopedia of Genes and Genomes （KEGG） pathway enrichment analysis and mapping were performed using the DAVID database and the Microbioinformatics platform. AutoDock 1.5.7 software was used to perform molecular docking between the top five active components and core targets. An animal model of simple obesity was established by feeding C57BL/6J mice a high-fat diet. The mice were administered ASR （2.06 g·kg^-1）， PR （2.06 g·kg^-1）， or ASR-PR （4.11 g·kg^-1） for 10 weeks， while the model group received an equal volume of purified water by gavage. After the administration period， the mice were sacrificed to measure body fat weight and serum levels of total cholesterol （TC）， triglycerides （TG）， high-density lipoprotein （HDL）， and low-density lipoprotein （LDL）. Hematoxylin-eosin （HE） staining was used to observe histopathological sections of liver and adipose tissue. Serum levels of leptin， interleukin-6 （IL-6）， and tumor necrosis factor-α （TNF-α） were determined by enzyme-linked immunosorbent assay （ELISA）， and the mRNA expression levels of epidermal growth factor receptor （EGFR） and signal transducer and activator of transcription 3 （STAT3） in liver tissue were detected by real-time quantitative polymerase chain reaction （Real-time PCR）. ResultsNetwork pharmacology and molecular docking results indicated that the treatment of simple obesity by ASR-PR may involve the regulation of protein expression of core targets EGFR and STAT3 by its main components MOL009760 （Siberian glycoside A_qt）， MOL003889 （methyl protodioscin_qt）， MOL009766 （resveratrol）， MOL006331 （4′，5-dihydroxyflavone）， and MOL004941 （baicalin）， thereby modulating the PI3K/Akt and JAK/STAT signaling pathways. The animal experiment results showed that compared with the normal group， the model group had significantly increased body weight， body fat weight， and serum levels of TG， TC， TNF-α， IL-6， and leptin （P<0.01）. EGFR mRNA expression was significantly elevated （P<0.05）， while STAT3 mRNA expression was significantly decreased （P<0.01）. Histological analysis revealed disordered hepatic architecture in the model group， with pronounced lipid vacuoles， cytoplasmic loosening， lipid accumulation， and steatosis. Adipocytes in white adipose tissue （WAT） and brown adipose tissue （BAT） of the model group exhibited markedly increased diameters， reduced cell counts per unit area， and irregular morphology. Compared with the model group， the ASR-PR group significantly reduced body weight， body fat weight， serum TC， IL-6， TNF-α， leptin levels， and EGFR mRNA expression （P<0.01）. TG levels were also significantly decreased （P<0.05）， while STAT3 mRNA expression was significantly increased （P<0.01）. Histopathological improvements included reduced size and number of hepatic lipid vacuoles and restoration of liver cell morphology toward that of the normal group. The diameter of adipocytes significantly decreased， and the number of adipocytes per unit area increased. ConclusionASR-PR may regulate the expression of key target proteins such as EGFR and STAT3 via its core active components， modulate the PI3K/Akt and JAK/STAT signaling pathways， repair damaged liver and adipose tissues， and thereby alleviate the progression of obesity in mice.

ObjectiveTo preliminarily explore the active components and target pathways of Angelicae Sinensis Radix-Polygonati Rhizoma （ASR-PR） herb pair in the treatment of simple obesity through network pharmacology and molecular docking， and to verify and investigate its mechanism of action via animal experiments. MethodsThe chemical constituents and targets of ASR and PR were predicted using the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform （TCMSP）. Targets related to simple obesity were identified by retrieving the GeneCards， Online Mendelian Inheritance in Man （OMIM）， Pharmacogenomics Knowledgebase （PharmGKB）， and DisGeNET databases. The intersection of drug and disease targets was used to construct an active component-target network using Cytoscape software. This network was imported into the STRING database to construct a protein-protein interaction （PPI） network， and topological analysis was conducted to identify core genes. Gene Ontology （GO） analysis and Kyoto Encyclopedia of Genes and Genomes （KEGG） pathway enrichment analysis and mapping were performed using the DAVID database and the Microbioinformatics platform. AutoDock 1.5.7 software was used to perform molecular docking between the top five active components and core targets. An animal model of simple obesity was established by feeding C57BL/6J mice a high-fat diet. The mice were administered ASR （2.06 g·kg^-1）， PR （2.06 g·kg^-1）， or ASR-PR （4.11 g·kg^-1） for 10 weeks， while the model group received an equal volume of purified water by gavage. After the administration period， the mice were sacrificed to measure body fat weight and serum levels of total cholesterol （TC）， triglycerides （TG）， high-density lipoprotein （HDL）， and low-density lipoprotein （LDL）. Hematoxylin-eosin （HE） staining was used to observe histopathological sections of liver and adipose tissue. Serum levels of leptin， interleukin-6 （IL-6）， and tumor necrosis factor-α （TNF-α） were determined by enzyme-linked immunosorbent assay （ELISA）， and the mRNA expression levels of epidermal growth factor receptor （EGFR） and signal transducer and activator of transcription 3 （STAT3） in liver tissue were detected by real-time quantitative polymerase chain reaction （Real-time PCR）. ResultsNetwork pharmacology and molecular docking results indicated that the treatment of simple obesity by ASR-PR may involve the regulation of protein expression of core targets EGFR and STAT3 by its main components MOL009760 （Siberian glycoside A_qt）， MOL003889 （methyl protodioscin_qt）， MOL009766 （resveratrol）， MOL006331 （4′，5-dihydroxyflavone）， and MOL004941 （baicalin）， thereby modulating the PI3K/Akt and JAK/STAT signaling pathways. The animal experiment results showed that compared with the normal group， the model group had significantly increased body weight， body fat weight， and serum levels of TG， TC， TNF-α， IL-6， and leptin （P<0.01）. EGFR mRNA expression was significantly elevated （P<0.05）， while STAT3 mRNA expression was significantly decreased （P<0.01）. Histological analysis revealed disordered hepatic architecture in the model group， with pronounced lipid vacuoles， cytoplasmic loosening， lipid accumulation， and steatosis. Adipocytes in white adipose tissue （WAT） and brown adipose tissue （BAT） of the model group exhibited markedly increased diameters， reduced cell counts per unit area， and irregular morphology. Compared with the model group， the ASR-PR group significantly reduced body weight， body fat weight， serum TC， IL-6， TNF-α， leptin levels， and EGFR mRNA expression （P<0.01）. TG levels were also significantly decreased （P<0.05）， while STAT3 mRNA expression was significantly increased （P<0.01）. Histopathological improvements included reduced size and number of hepatic lipid vacuoles and restoration of liver cell morphology toward that of the normal group. The diameter of adipocytes significantly decreased， and the number of adipocytes per unit area increased. ConclusionASR-PR may regulate the expression of key target proteins such as EGFR and STAT3 via its core active components， modulate the PI3K/Akt and JAK/STAT signaling pathways， repair damaged liver and adipose tissues， and thereby alleviate the progression of obesity in mice.

BACKGROUND: The available evidence regarding the benefits of percutaneous coronary intervention (PCI) on patients receiving dialysis with coronary artery disease (CAD) is limited and inconsistent. This study aimed to evaluate the association between PCI and clinical outcomes as compared with medical therapy alone in patients undergoing dialysis with CAD in China. METHODS: This multicenter, retrospective study was conducted in 30 tertiary medical centers across 12 provinces in China from January 2015 to June 2021 to include patients on dialysis with CAD. The primary outcome was major adverse cardiovascular events (MACE), defined as a composite of cardiovascular death, non-fatal myocardial infarction, and non-fatal stroke. Secondary outcomes included all-cause death, the individual components of MACE, and Bleeding Academic Research Consortium criteria types 2, 3, or 5 bleeding. Multivariable Cox proportional hazard models were used to assess the association between PCI and outcomes. Inverse probability of treatment weighting (IPTW) and propensity score matching (PSM) were performed to account for potential between-group differences. RESULTS: Of the 1146 patients on dialysis with significant CAD, 821 (71.6%) underwent PCI. After a median follow-up of 23.0 months, PCI was associated with a 43.0% significantly lower risk for MACE (33.9% [ n  = 278] vs . 43.7% [ n  = 142]; adjusted hazards ratio 0.57, 95% confidence interval 0.45-0.71), along with a slightly increased risk for bleeding outcomes that did not reach statistical significance (11.1% vs . 8.3%; adjusted hazards ratio 1.31, 95% confidence interval, 0.82-2.11). Furthermore, PCI was associated with a significant reduction in all-cause and cardiovascular mortalities. Subgroup analysis did not modify the association of PCI with patient outcomes. These primary findings were consistent across IPTW, PSM, and competing risk analyses. CONCLUSION This study indicated that PCI in patients on dialysis with CAD was significantly associated with lower MACE and mortality when comparing with those with medical therapy alone, albeit with a slightly increased risk for bleeding events that did not reach statistical significance.
Humans ; Percutaneous Coronary Intervention/methods* ; Male ; Female ; Coronary Artery Disease/drug therapy* ; Retrospective Studies ; Renal Dialysis/methods* ; Middle Aged ; Aged ; China ; Proportional Hazards Models ; Treatment Outcome

Objective To explore the application of plasma 7 microRNA (miR7) testing in the differential diagnosis of very early-stage hepatocellular carcinoma (HCC). Methods This study is a multicenter real-world study. Patients with single hepatic lesion (maximum diameter≤2 cm) who underwent plasma miR7 testing at Zhongshan Hospital, Fudan University, Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, Anhui Provincial Hospital, and Peking University People’s Hospital between January 2019 and December 2024 were retrospectively enrolled. Patients were divided into very early-stage HCC group and non-HCC group, and the clinical pathological characteristics of the two groups were compared. The value of plasma miR7 levels, alpha-fetoprotein (AFP), and des-gamma-carboxy prothrombin (DCP) in the differential diagnosis of very early-stage HCC was evaluated using receiver operating characteristic (ROC) curves and area under the curve (AUC). In patients with both negative AFP and DCP (AFP＜20 ng/mL, DCP＜40 mAU/mL), the diagnostic value of plasma miR7 for very early-stage HCC was analyzed. Results A total of 64 528 patients from 4 hospitals underwent miR7 testing, and 1 682 were finally included, of which 1 073 were diagnosed with very early-stage HCC and 609 were diagnosed with non-HCC. The positive rate of miR7 in HCC patients was significantly higher than that in non-HCC patients (67.9% vs 24.3%, P＜0.001). ROC curves showed that the AUCs for miR7, AFP, and DCP in distinguishing HCC patients from the non-HCC individuals were 0.718, 0.682, and 0.642, respectively. The sensitivities were 67.85%, 43.71%, and 44.45%, and the specificities were 75.70%, 92.78%, and 83.91%, respectively. The pairwise comparison of AUCs showed that the diagnostic efficacy of plasma miR7 detection was significantly better than that of AFP or DCP (P＜0.05). Although its specificity was slightly lower than AFP and DCP, the sensitivity was significantly higher. Among patients negative for both AFP and DCP, miR7 maintained an AUC of 0.728 for diagnosing very early-stage HCC, with 67.82% sensitivity and 77.73% specificity. Conclusions Plasma miR7 testing is a potential molecular marker with high sensitivity and specificity for the differential diagnosis of small hepatic nodules. In patients with very early-stage HCC lacking effective molecular markers (negative for both AFP and DCP), miR7 can serve as a novel and effective molecular marker to assist diagnosis.

ObjectiveTo investigate the mechanisms of mitochondrial dynamics and metabolic reprogramming in the treatment of diabetic nephropathy （DN） by Shenxiao Tongluo prescription via the peroxisome proliferator-activated receptor γ coactivator-1α （PGC-1α）/sirtuin-3 （SIRT3）/hypoxia-inducible factor-1α （HIF-1α） signaling pathway. MethodsSixty-five SD rats were randomized into a sham group （10 rats） and a modeling group （55 rats）， and the modeling rats underwent left nephrectomy and intraperitoneal injection of streptozotocin （35 mg·kg^-1） to prepare a DN model. After successful modeling， the rats were randomized into model， empagliflozin （10 mg·kg^-1）， and low-， medium-， and high-dose （7.656， 15.312， 30.624 g·kg^-1， respectively） Shenxiao Tongluo prescription groups. The urine microalbumin （UmAlb）， blood urea nitrogen （BUN）， and serum creatinine （SCr） levels of rats in each group were assessed after continuous gavage for 8 weeks. The corresponding kits were used to measure the levels of lactate， superoxide dismutase （SOD）， and malondialdehyde （MDA） in the kidney tissue. Hematoxylin-eosin staining， Masson staining， and periodic acid-Schiff staining were performed to observe the pathological changes in the kidney tissue. Transmission electron microscopy was employed to observe mitochondrial morphology. Immunohistochemistry was employed to determine the expression levels of dynamin-related protein 1 （DRP1） and pyruvate kinase M2 （PKM2） in the kidney tissue. Western blot was adopted to assess the protein levels of PGC-1α， SIRT3， HIF-1α， dynamin-related protein 1 （Drp1）， optic atrophy 1 （OPA1）， hexokinase 2 （HK2）， and pyruvate kinase M2 （PKM2） in the kidney tissue. ResultsCompared with the sham group， the model group showed elevated levels of UmAlb， BUN， SCr， lactate， and MDA， decreased SOD level （P<0.05）， glomerular hypertrophy， thickening of the mesangial basement membrane， vacuolar degeneration of renal tubular epithelial cells， and infiltration of renal interstitial inflammatory cells， oval mitochondria with disordered， blurred or disappearing cristae， down-regulated protein levels of PGC-1α， SIRT3， and OPA1， and up-regulated protein levels of HIF-1α， DRP1， HK2， and PKM2 （P<0.05）. Compared with the model group， the treatment in all the groups increased the body weight， lowered the levels of GLU， UmAlb， BUN， and MDA， raised the level of SOD， alleviated the pathological damage in the kidney tissue and mitochondrial damage， up-regulated the expression of PGC-1α， SIRT3， and OPA1， and down-regulated the expression of HIF-1α， DRP1， and PKM2 （P<0.05）. Empagliflozin and Shenxiao Tongluo prescription at medium and high doses lowered the levels of SCr and lactate and down-regulated the expression of HK2 （P<0.05）， which had no statistical significance in the low-dose Shenxiao Tongluo prescription group. ConclusionShenxiao Tongluo prescription may regulate mitochondrial dynamics and metabolic reprogramming by activating the PGC-1α/SIRT3/HIF-1α pathway， thereby alleviating oxidative damage in the kidney tissue and delaying the progression of DN.

ObjectiveTo investigate the mechanisms of mitochondrial dynamics and metabolic reprogramming in the treatment of diabetic nephropathy （DN） by Shenxiao Tongluo prescription via the peroxisome proliferator-activated receptor γ coactivator-1α （PGC-1α）/sirtuin-3 （SIRT3）/hypoxia-inducible factor-1α （HIF-1α） signaling pathway. MethodsSixty-five SD rats were randomized into a sham group （10 rats） and a modeling group （55 rats）， and the modeling rats underwent left nephrectomy and intraperitoneal injection of streptozotocin （35 mg·kg^-1） to prepare a DN model. After successful modeling， the rats were randomized into model， empagliflozin （10 mg·kg^-1）， and low-， medium-， and high-dose （7.656， 15.312， 30.624 g·kg^-1， respectively） Shenxiao Tongluo prescription groups. The urine microalbumin （UmAlb）， blood urea nitrogen （BUN）， and serum creatinine （SCr） levels of rats in each group were assessed after continuous gavage for 8 weeks. The corresponding kits were used to measure the levels of lactate， superoxide dismutase （SOD）， and malondialdehyde （MDA） in the kidney tissue. Hematoxylin-eosin staining， Masson staining， and periodic acid-Schiff staining were performed to observe the pathological changes in the kidney tissue. Transmission electron microscopy was employed to observe mitochondrial morphology. Immunohistochemistry was employed to determine the expression levels of dynamin-related protein 1 （DRP1） and pyruvate kinase M2 （PKM2） in the kidney tissue. Western blot was adopted to assess the protein levels of PGC-1α， SIRT3， HIF-1α， dynamin-related protein 1 （Drp1）， optic atrophy 1 （OPA1）， hexokinase 2 （HK2）， and pyruvate kinase M2 （PKM2） in the kidney tissue. ResultsCompared with the sham group， the model group showed elevated levels of UmAlb， BUN， SCr， lactate， and MDA， decreased SOD level （P<0.05）， glomerular hypertrophy， thickening of the mesangial basement membrane， vacuolar degeneration of renal tubular epithelial cells， and infiltration of renal interstitial inflammatory cells， oval mitochondria with disordered， blurred or disappearing cristae， down-regulated protein levels of PGC-1α， SIRT3， and OPA1， and up-regulated protein levels of HIF-1α， DRP1， HK2， and PKM2 （P<0.05）. Compared with the model group， the treatment in all the groups increased the body weight， lowered the levels of GLU， UmAlb， BUN， and MDA， raised the level of SOD， alleviated the pathological damage in the kidney tissue and mitochondrial damage， up-regulated the expression of PGC-1α， SIRT3， and OPA1， and down-regulated the expression of HIF-1α， DRP1， and PKM2 （P<0.05）. Empagliflozin and Shenxiao Tongluo prescription at medium and high doses lowered the levels of SCr and lactate and down-regulated the expression of HK2 （P<0.05）， which had no statistical significance in the low-dose Shenxiao Tongluo prescription group. ConclusionShenxiao Tongluo prescription may regulate mitochondrial dynamics and metabolic reprogramming by activating the PGC-1α/SIRT3/HIF-1α pathway， thereby alleviating oxidative damage in the kidney tissue and delaying the progression of DN.

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.