Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy最新文献

{"title":"Zinc-mediated rigid cyclization enables highly enantioselective fluorescent recognition of chiral amino alcohols in aqueous phase.","authors":"Chunmei Chen, Shijun Chen, Xianzhe Yu, Qiye Liu, Cailing Fan, Weijie Chi, Chaoyuan Zeng","doi":"10.1016/j.saa.2026.128892","DOIUrl":"10.1016/j.saa.2026.128892","url":null,"abstract":"<p><p>Chiral amino alcohols represent a crucial class of chiral amines with significant applications across various fields. Herein, a novel BINOL-based fluorescent probe(R)-5, featuring dual aldehyde and hydroxyl functionalities, was rationally designed and synthesized. (R)-5 exhibited excellent enantioselective recognition of D-valinol over L-valinol in aqueous media, with a detection limit (LOD) of 12.75 μM for D-valinol and an enantioselective fluorescence enhancement ratio [ef = (I_D - I₀)/(I_L-I₀)] of 20.50. Comprehensive structural characterization suggests that the observed fluorescence enhancement is attributed to the formation of a rigid cyclic coordination complex, facilitated by coordination between the imine and Zn²⁺. This work presents a novel and efficient strategy for the simultaneous determination of the enantiomeric composition and concentration of chiral amino alcohols.</p>","PeriodicalId":94213,"journal":{"name":"Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy","volume":"366 ","pages":"128892"},"PeriodicalIF":4.6,"publicationDate":"2027-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149083538","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Machine learning algorithms empower traditional Chinese medicine: Breakthrough approaches from chemical component analysis to pharmacological mechanism research.","authors":"Zheyue Han, Fuxia Zhao, Haonan Xu, Zhenkai Zhang, Jiatong Wu, Yiqiang Zhang, Juan Pan, Mengmeng Li, Zhichao Hao, Wei Guan, Peng Jiang, Yumeng Luo, Xiaochi Ma, Haixue Kuang, Yan Liu","doi":"10.1016/j.saa.2026.128859","DOIUrl":"10.1016/j.saa.2026.128859","url":null,"abstract":"<p><p>Traditional Chinese medicine studies are constrained by complex chemical components and multi-target biological interactions, making conventional analytical methods inefficient for mechanism exploration. As a data-driven technique, machine learning has been widely adopted to tackle diverse complex challenges in TCM research. Nevertheless, most existing reviews merely summarize individual algorithms or scattered application cases, lacking systematic task-based classification and unified operational standards. Based on literature published from 2022 to 2026, this review systematically classifies machine learning applications in TCM into four core research tasks. A task-oriented framework is constructed to match different algorithms with appropriate TCM research scenarios. This paper further establishes a complete standardized workflow covering algorithm selection, model construction and model evaluation to standardize ML applications in TCM research. This review also summarizes the applications of machine learning in TCM quality control, bioactive compound identification, processing analysis, and mechanism elucidation. Furthermore, major bottlenecks in current ML-based TCM studies are analyzed, and feasible optimization strategies are proposed. This review aims to provide clear and practical guidelines for the standardized application of machine learning in future TCM research.</p>","PeriodicalId":94213,"journal":{"name":"Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy","volume":"366 ","pages":"128859"},"PeriodicalIF":4.6,"publicationDate":"2027-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149121055","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multimodal Fourier transform infrared spectroscopy and second- and third-harmonic generation imaging for integrated characterization of oral squamous cell carcinoma.","authors":"Kamila Wolnica, Mateusz Dulski, Katarzyna Merkel, Robert Gawecki, Daria Wziątek-Kuczmik, Ievgeniia Iermak, Jacek Pająk, Iwona Niedzielska","doi":"10.1016/j.saa.2026.128884","DOIUrl":"10.1016/j.saa.2026.128884","url":null,"abstract":"<p><p>Fourier Transform Infrared (FTIR) spectroscopy combined with nonlinear optical imaging was employed to characterize molecular and structural alterations associated with oral squamous cell carcinoma (OSCC) progression in control, margin, and tumor tissues. Spectral and imaging datasets were evaluated using multivariate statistical analysis and receiver operating characteristic (ROC) analysis to identify disease-associated features and assess their discriminatory performance. Integration of second- and third-harmonic generation (SHG/THG) imaging with histopathological staining provided complementary information on tissue architecture, with SHG contrast reflecting extracellular matrix and collagen organization and THG contrast highlighting cellular and nuclear features. FTIR analysis revealed coordinated biochemical alterations during OSCC progression, including reduced α-helical protein content with preservation of β-sheet features, lipid remodeling characterized by decreased ester carbonyl content, lower CH₂/CH₃ ratios, and reduced lipid unsaturation, altered glycogen-related signals, and nucleic acid reorganization characterized by decreased DNA and increased RNA contributions. Although many spectral parameters showed statistically significant differences, their individual diagnostic performance was limited, particularly by low specificity. A selected set of spectral parameters, including the Amide I / Ester carbonyl ratio, α-helix content, α-helix/β-sheet ratio, and nucleic acid-associated markers, showed relatively higher discriminatory performance, with AUC values of 0.80-0.93, sensitivities of 66-93%, and specificities of 1-28%. These findings demonstrate that FTIR spectroscopy captures distinct molecular alterations associated with OSCC progression, while SHG/THG imaging provides complementary structural information. Their integration therefore offers a multimodal, label-free framework for spectro-histopathological characterization of OSCC and supports the use of multiple biochemical and structural descriptors rather than individual spectral markers.</p>","PeriodicalId":94213,"journal":{"name":"Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy","volume":"366 ","pages":"128884"},"PeriodicalIF":4.6,"publicationDate":"2027-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149121126","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Raman a banana: mapping banana peel composition and ripening process by Raman spectroscopy.","authors":"Zdeněk Pilát, Pavlína Modlitbová, Markéta Benešová, Martin Šiler, Silvie Bernatová, Jan Ježek, Filip Plešinger, Ondřej Doskočil, Ota Samek, Pavel Zemánek","doi":"10.1016/j.saa.2026.128898","DOIUrl":"10.1016/j.saa.2026.128898","url":null,"abstract":"<p><p>Banana peel was analyzed by Raman microspectroscopy and its chemical components were identified by comparison of the acquired spectra with reference substances, including plant pigments, pectin, and flavonoids. Compounds of pharmaceutical and food-industry relevance - chrysin, lutein, and pectin - were detected. Raman mapping revealed microscopic inclusions of calcium oxalate in the epidermis, the size of which increased during the ripening process. The progress of ripening was associated with changing relative intensities of the Raman signals of chlorophyll, carotenoids, and further banana peel constituents: over the whole ripening period the intensity of the chlorophyll band at 1328 cm^-1 decreased 2.7-fold while that of the carotenoid band at 1527 cm^-1 increased 1.7-fold. A normalized difference index derived from the two bands increased monotonically with storage time and provides a self-referencing measure of ripening independent of the absolute instrument response. Standardized colorimetric analysis of the same fruits confirmed the ripeness classification by independent instrumental means. The results shed light on important aspects of the banana ripening process and demonstrate the usefulness of Raman spectroscopy for qualitative and semi-quantitative, non-contact and non-destructive analysis of banana peel composition. They further provide a reference for Raman spectroscopic evaluation of growth parameters as an extension of hyperspectral or multimodal imaging, and support the consideration of banana peel as a renewable, eco-friendly source of valuable compounds.</p>","PeriodicalId":94213,"journal":{"name":"Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy","volume":"366 ","pages":"128898"},"PeriodicalIF":4.6,"publicationDate":"2027-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149121262","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Chemometrics-assisted FTIR and machine learning for biochemical profiling of human skeletal muscles in relation to type 2 diabetes.","authors":"Mohamed Elwy Abdelhamed Abdelmonaem, Chiedozie Kenneth Ugwoke, Armin Alibegović, Erika Cvetko, Nejc Umek, Jože Grdadolnik, Barbara Zupančič","doi":"10.1016/j.saa.2026.128893","DOIUrl":"10.1016/j.saa.2026.128893","url":null,"abstract":"<p><p>Skeletal muscle is the major site of insulin-stimulated glucose uptake, making it highly susceptible to metabolic disorders. This study investigates Type 2 diabetes mellitus-associated alterations in the macromolecular composition of functionally distinct human skeletal muscles. These modifications are crucial to understanding the molecular basis of diabetic myopathies but remain underexplored. Five skeletal muscles from 80 individuals (40 controls and 40 individuals with Type 2 diabetes) were analysed. To overcome the limitations of traditional histochemical assays, we employed infrared spectroscopy combined with a chemometric decomposition approach to resolve spectra into five spectral components, each dominated by a specific macromolecular group. The corresponding component weights quantified and compared relative macromolecular proportions between the diabetic and control groups for different muscle types. Compared with controls, muscles from individuals with diabetes showed a consistent pattern of higher lipid levels and lower glycogen levels, with a statistically significant increase in lipids within the diaphragm. This biochemical profiling was enriched by building machine learning classifiers, which achieved low-to-modest but statistically significant AUC performance of up to 72% (p < 0.05) in discriminating diabetic from healthy muscle. Our work demonstrates an efficient methodology for the simultaneous screening of various macromolecular constituents using spectral data obtained in a single experiment. It provides medically relevant insights into the biological patterns of diabetes-associated skeletal muscle remodelling to guide future research. However, an independent validation should be performed to overcome this study limitation and draw definitive conclusions.</p>","PeriodicalId":94213,"journal":{"name":"Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy","volume":"366 ","pages":"128893"},"PeriodicalIF":4.6,"publicationDate":"2027-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149121245","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Corrigendum to \"Surface-enhanced Raman spectroscopy of serum exosomes coupled with support vector machine for diagnosis of Parkinson's disease\" [Spectrochim. Acta Part A: Molecul. Biomolecul. Spectrosc. 353 (2026) 127573].","authors":"Xinran Liu, Xinming Wei, Xiangxiang Zheng, Liang Xu, Guohua Wu, Keke Feng","doi":"10.1016/j.saa.2026.128860","DOIUrl":"10.1016/j.saa.2026.128860","url":null,"abstract":"","PeriodicalId":94213,"journal":{"name":"Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy","volume":"366 ","pages":"128860"},"PeriodicalIF":4.6,"publicationDate":"2027-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149121043","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multimodal deep learning-driven Raman spectroscopy for rapid identification of closely related pathogenic Bacillus species.","authors":"Jiazheng Sun, Haodong Liu, Kaier Yang, Xiangyun Ma, Qifeng Li","doi":"10.1016/j.saa.2026.128883","DOIUrl":"10.1016/j.saa.2026.128883","url":null,"abstract":"<p><p>Rapid and fine-grained identification of closely related pathogenic Bacillus species is a critical challenge in optical biosensing and biomolecular spectroscopy, hindered by their highly similar biochemical compositions and severe spectral overlap. Existing single-modality spectroscopic analysis and computational frameworks frequently fail to resolve these subtle spectral shifts, particularly when confronted with instrumental noise, background interference, and limited training data. To overcome these spectroscopic bottlenecks, this study proposes the Spectral Continuous Wavelet Transform Transformer Network (Spec-CWT-TransNet), a novel multimodal deep learning framework designed for the precise identification of closely related microbial strains. By transforming 1D Raman spectra into 2D wavenumber-scale scalograms utilizing the Continuous Wavelet Transform (CWT), the proposed method synergizes the robust local topological texture extraction of Convolutional Neural Networks (CNNs) with the global dependency modeling capabilities of Transformers. Furthermore, a Fisher Score-based scale optimization strategy is introduced to adaptively select discriminative frequency bands, effectively suppressing redundant instrumental and background noise. Experimental results demonstrate that Spec-CWT-TransNet achieves a superior overall identification accuracy of 99.38% on four Bacillus species. Notably, under limited data conditions (using only 40% of training data), the framework maintains a robust accuracy of 92.50%, quantifiably surpassing the performance of traditional 1D baselines trained on the full dataset. Moreover, under severe spectral interference, the model retains 90.63% accuracy, outperforming unimodal baseline models by margins exceeding 22.5%. Overall, this multimodal Raman data analysis methodology demonstrates great potential for rapid, robust, and reproducible spectral discrimination in non-destructive optical sensing and bioanalytical applications.</p>","PeriodicalId":94213,"journal":{"name":"Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy","volume":"366 ","pages":"128883"},"PeriodicalIF":4.6,"publicationDate":"2027-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149083470","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development of a europium-thiosemicarbazide-based nano-optical biosensor for early detection of HE4 in ovarian cancer.","authors":"Ahmed Wanas, Ameera S Alshinnawy, Moshira M Abdel Waled, Mohamed R Mohamed, Faisal K Algethami, Mohamed S Attia, Mona N Abou-Omar","doi":"10.1016/j.saa.2026.128891","DOIUrl":"10.1016/j.saa.2026.128891","url":null,"abstract":"&lt;p&gt;&lt;p&gt;Early diagnosis of ovarian cancer remains a major clinical challenge. Human epididymis protein 4 (HE4) has emerged as a promising biomarker for early OC detection. To address limitations of existing HE4 quantification methods, we developed a nano-optical biosensor based on a luminescent europium-thiosemicarbazide (Eu-TSC) complex immobilized within a modified epoxy-cellulose thin-film matrix. The Eu-TSC thin film exhibited characteristic europium excitation and emission bands at 395 nm and 617 nm. Following functionalization with a monoclonal anti-HE4 antibody, the biosensor displayed a concentration-dependent fluorescence quenching response upon HE4 binding. A linear analytical response was obtained over 0.05-0.37 ng mL&lt;sup&gt;-1&lt;/sup&gt;, with LOD of 0.108 ng mL&lt;sup&gt;-1&lt;/sup&gt;. However, the LOQ (0.327 ng mL&lt;sup&gt;-1&lt;/sup&gt;) lies close to the upper limit of the linear range, the practical validated quantification interval is extremely narrow (approximately 0.327-0.37 ng mL&lt;sup&gt;-1&lt;/sup&gt;). This interval is far too narrow to cover the clinical HE4 range, and the present sensor should therefore be regarded as a preliminary proof-of-concept platform rather than a clinically suitable quantitative assay. Time-resolved luminescence measurements are consistent with a quenching-based sensing response, but the present experiments do not distinguish between FRET, PET, or other quenching pathways. Molecular docking and 50 ns molecular dynamics simulations provide a computational rationale for ligand selection and suggest possible surface binding of thiosemicarbazide-based ligands to HE4. Selectivity studies against twelve potential interferents revealed relative responses below 5% for all tested analytes when tested individually at clinically relevant concentrations (n = 3 per interferent; one-way ANOVA with Dunnett's test). However, because interferents were tested individually rather than as pooled mixtures, and because cross-reactivity with structurally related proteins cannot be fully excluded, these selectivity results are preliminary and do not establish excellent selectivity in complex serum. The biosensor was further evaluated using 21 human serum samples (11 OC patients, 10 healthy controls), demonstrating agreement with a commercial HE4 ELISA assay (R&lt;sup&gt;2&lt;/sup&gt; &gt; 0.98). Recovery values were re-verified for all samples; several values fall outside the commonly accepted 85-115% range and are flagged in Tables 1A and B. Therefore, the analytical accuracy of the present method is preliminary and not satisfactory for clinical quantification. ROC analysis yielded an AUC of 1.000, with 90.9% sensitivity and 100% specificity. These findings are based on a limited case-control cohort (n = 21) and must be interpreted as a preliminary proof-of-concept analytical and clinical evaluation, not as a clinically validated diagnostic platform. Larger prospective, multi-center studies including patients with benign gynecological conditions are required before clinical i","PeriodicalId":94213,"journal":{"name":"Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy","volume":"366 ","pages":"128891"},"PeriodicalIF":4.6,"publicationDate":"2027-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149083360","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Sorption-spectroscopic determination of chromium (III) ions using eriochrome black T immobilized on silk fibroin fiber.","authors":"Zulaykho Smanova, Khilola Usmanova, Muattar Mamedova, Ayjamal Turambetova, Nazira Madusmanova, Muborak Abdullayeva, Ulugbek Akhmadjonov, Go'zal Yunusova, Feruza Tojiboyeva, Mavjuda Rakhmatullaeva, Dilrabo Muhammadiyeva, Enrico Benassi","doi":"10.1016/j.saa.2026.128464","DOIUrl":"10.1016/j.saa.2026.128464","url":null,"abstract":"<p><strong>Background: </strong>Chromium (III) ions are widely present in industrial and environmental waters, and their reliable determination is essential for environmental monitoring and pollution control. Conventional spectrophotometric methods often suffer from limited sensitivity, selectivity, and stability. To overcome these limitations, a sorption-spectrophotometric approach based on reagent immobilization was developed. In this study, Eriochrome Black T (EBT) was immobilized on silk fibroin fiber to enhance analytical performance in the determination of Cr(III) ions.</p><p><strong>Results: </strong>Immobilization of EBT significantly improved the stability, sensitivity, and selectivity of the method. Formation of the Fib-EBT-Cr(III) system was confirmed by FTIR spectroscopy (CrO band at 607 cm^-1), X-ray fluorescence analysis, and SEM-EDS, which indicated 18.00 ± 0.01 wt% chromium in the immobilized complex. Optimal analytical conditions were pH 8, λ = 618 nm, and a universal buffer. The method showed a detection limit of 1.2 μg/L, linearity for Cr(III) amounts in the range of 5-50 μg (R² = 0.9991), and an analysis time of 10 min. Interference from Al³⁺, Co²⁺, and Cd²⁺ was effectively masked, while Ni²⁺, Fe³⁺, and Cu²⁺ caused no interference. Validation using industrial wastewater samples demonstrated relative errors ≤6% and Sr ≤ 0.085. Quantum chemical calculations identified the most stable isomer of EBT (rotamer B) and its complex with Cr(III) in aqueous solution. Comparison of computed and experimental IR spectra confirmed the dominance of rotamer B, the involvement of the azo group in coordination, and the formation of a CrO bond (experimental 607 cm^-1; computed 600 cm^-1). Complexation was found to induce trans-to-cis isomerization of EBT, a phenomenon not previously reported in the literature.</p><p><strong>Significance: </strong>The developed sorption-spectrophotometric method provides a rapid, economical, and accurate approach for the determination of Cr(III) ions in environmental waters. Its high sensitivity, good selectivity, and successful application to real industrial wastewater samples indicate strong potential for routine laboratory analysis as well as field monitoring of chromium contamination.</p>","PeriodicalId":94213,"journal":{"name":"Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy","volume":"363 Pt 2","pages":"128464"},"PeriodicalIF":4.6,"publicationDate":"2026-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148521437","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Raman spectroscopy combined with machine learning algorithms for discrimination of cancer cell death pathways: Apoptosis and pyroptosis.","authors":"Wen Lei, Juan Li, Fengling Li, Ayiben Wenhua, Xining Yang, Bingqian Hou, Jingjing Xia, Lei Zhang","doi":"10.1016/j.saa.2026.128470","DOIUrl":"10.1016/j.saa.2026.128470","url":null,"abstract":"<p><strong>Background and objectives: </strong>Identifying drug-induced cancer cell death pathways is crucial for understanding the mechanisms of drug action. However, traditional biological experimental methods are limited by their time-consuming nature and high cost. This aim of this study is to develop classification models that can predict apoptosis and pyroptosis in cancer cells, combining Raman spectroscopy with machine learning algorithms.</p><p><strong>Methods: </strong>Cisplatin and dronedarone, two drugs known to induce distinct cell death pathways, were selected to validate the predictive performance of the constructed models. Additionally, Gradient-weighted Class Activation Mapping (Grad-CAM) visualization method was employed to identify key feature peaks. Model predictions were further validated through cellular morphological observations and Western blot analysis.</p><p><strong>Results: </strong>Both algorithms achieved prediction accuracies exceeding 90% for apoptosis and pyroptosis. Predictions of cell death pathways induced by cisplatin and dronedarone were concentrated on apoptosis and pyroptosis, respectively. Visualization results using gradient-weighted class activation mapping showed that apoptosis is associated with characteristic peaks of phenylalanine, while pyroptosis is associated with characteristic peaks of tyrosine. These results are consistent with existing research findings.</p><p><strong>Conclusion: </strong>This study demonstrates that Raman spectroscopy combined with machine learning algorithms can accurately predict cancer cell death pathways, and that visualization techniques can be used to identify biomarkers associated with these pathways, providing reliable technical support for research into drug mechanisms of action.</p>","PeriodicalId":94213,"journal":{"name":"Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy","volume":"363 Pt 2","pages":"128470"},"PeriodicalIF":4.6,"publicationDate":"2026-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148581898","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}