Pub Date : 2026-04-01Epub Date: 2025-12-10DOI: 10.1016/j.compbiolchem.2025.108826
Hetvi Shah , Adikrishna Murali Mohan , Rushabh Shah , Drashti Mehta , A.V. Ramachandran , Parth Pandya
Glioblastoma multiforme (GB) is the most aggressive and lethal primary brain tumor, with limited biomarkers for diagnosis and therapeutic targeting. This study aimed to investigate the regulatory effects of Kisspeptin-10 on epithelial–mesenchymal transition (EMT) and apoptosis in GB by integrating transcriptomic profiling, network analysis, and in-vitro validation. Kisspeptin-10, a metastasis-suppressor peptide known to modulate EMT and apoptotic pathways in several cancers, has not been previously explored in GB. Differentially expressed genes (DEGs) were identified from publicly available GEO datasets using limma, followed by STRING-based protein–protein interaction (PPI) analysis, cytoHubba-based hub gene ranking, and construction of miRNA–mRNA regulatory networks. A total of 1401 DEGs were identified, including 859 upregulated and 542 downregulated genes, enriched in pathways associated with EMT regulation, cell-cycle progression, extracellular matrix remodeling, and apoptosis. Hub genes such as CDK1, CDC20, JUN, and FABP5 were identified, while miR-200, miR-345, and miR-577 emerged as key regulatory miRNAs linked to EMT and apoptotic signaling. In-vitro validation further supported the modulatory effects of Kisspeptin-10 on EMT and apoptosis markers in GB cells. These findings highlight the diagnostic and therapeutic relevance of Kisspeptin-10–associated molecular regulation in GB. This is the first study to integrate transcriptomics, miRNA–mRNA network analysis, and experimental validation to elucidate Kisspeptin-10–mediated modulation of GB progression.
{"title":"Integrated transcriptomics and miRNA-mRNA network analysis reveals Kisspeptin-10 mediated regulation of EMT and apoptosis in glioblastoma","authors":"Hetvi Shah , Adikrishna Murali Mohan , Rushabh Shah , Drashti Mehta , A.V. Ramachandran , Parth Pandya","doi":"10.1016/j.compbiolchem.2025.108826","DOIUrl":"10.1016/j.compbiolchem.2025.108826","url":null,"abstract":"<div><div>Glioblastoma multiforme (GB) is the most aggressive and lethal primary brain tumor, with limited biomarkers for diagnosis and therapeutic targeting. This study aimed to investigate the regulatory effects of Kisspeptin-10 on epithelial–mesenchymal transition (EMT) and apoptosis in GB by integrating transcriptomic profiling, network analysis, and in-vitro validation. Kisspeptin-10, a metastasis-suppressor peptide known to modulate EMT and apoptotic pathways in several cancers, has not been previously explored in GB. Differentially expressed genes (DEGs) were identified from publicly available GEO datasets using limma, followed by STRING-based protein–protein interaction (PPI) analysis, cytoHubba-based hub gene ranking, and construction of miRNA–mRNA regulatory networks. A total of 1401 DEGs were identified, including 859 upregulated and 542 downregulated genes, enriched in pathways associated with EMT regulation, cell-cycle progression, extracellular matrix remodeling, and apoptosis. Hub genes such as CDK1, CDC20, JUN, and FABP5 were identified, while miR-200, miR-345, and miR-577 emerged as key regulatory miRNAs linked to EMT and apoptotic signaling. In-vitro validation further supported the modulatory effects of Kisspeptin-10 on EMT and apoptosis markers in GB cells. These findings highlight the diagnostic and therapeutic relevance of Kisspeptin-10–associated molecular regulation in GB. This is the first study to integrate transcriptomics, miRNA–mRNA network analysis, and experimental validation to elucidate Kisspeptin-10–mediated modulation of GB progression.</div></div>","PeriodicalId":10616,"journal":{"name":"Computational Biology and Chemistry","volume":"121 ","pages":"Article 108826"},"PeriodicalIF":3.1,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145734064","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-04-01Epub Date: 2025-12-13DOI: 10.1016/j.compbiolchem.2025.108844
Wan-li Wang , Qiao Xiong , Bo Ma , Xin-hua Liang , Ya-ling Tang
Objective
This study aims to comprehensively characterize ACOX3 as a novel pan-cancer biomarker by assessing its expression heterogeneity, clinical relevance, tumor-immune interactions and therapeutic potential across multiple cancer types.
Methods
The multi-omics analyses of the ACOX3 expression pattern were performed. Prognostic and diagnostic significance was evaluated by Cox regression, Kaplan–Meier survival and ROC analyses. Immune correlates were assessed in terms of immune cell infiltration, checkpoint and immunomodulatory activity. Drug sensitivity was predicted through molecular docking and molecular dynamics simulations to evaluate binding affinity and complex stability. Experimental validation was conducted in HNSCC cell lines.
Results
ACOX3 was significantly upregulated in KICH, PRAD and THCA, and downregulated in COAD, HNSCC, KIRP, LIHC and STAD. The OS Cox regression showed high ACOX3 expression was associated with a favorable prognosis in HNSCC but poor outcomes in LGG and UVM. The ROC curves showed that the AUC for ESCA, GBM, OV, PAAD, STES and WT exceeded 0.8. ACOX3 expression positively correlated with CD4⁺T, CD8⁺T and NK cells in HNSCC. Single-cell and spatial transcriptomics revealed ACOX3 enrichment in malignant regions, particularly in CD4⁺T, CD8⁺T and CD8⁺Tex cells. Drug screening prioritized AZD6482 and TGX-221 as high-affinity ACOX3 inhibitors, and the AZD6482 showed stable binding in MD simulations. Functional experiments confirmed that ACOX3 overexpression suppressed HNSCC cell proliferation, invasion and migration.
Conclusion
ACOX3 represents a dual diagnostic and prognostic biomarker with broad pan-cancer relevance, exhibiting distinct immune correlates and therapeutic potential.
{"title":"Multi-omics profiling of ACOX3 unveils pan-cancer clinical biomarker potential","authors":"Wan-li Wang , Qiao Xiong , Bo Ma , Xin-hua Liang , Ya-ling Tang","doi":"10.1016/j.compbiolchem.2025.108844","DOIUrl":"10.1016/j.compbiolchem.2025.108844","url":null,"abstract":"<div><h3>Objective</h3><div>This study aims to comprehensively characterize ACOX3 as a novel pan-cancer biomarker by assessing its expression heterogeneity, clinical relevance, tumor-immune interactions and therapeutic potential across multiple cancer types.</div></div><div><h3>Methods</h3><div>The multi-omics analyses of the ACOX3 expression pattern were performed. Prognostic and diagnostic significance was evaluated by Cox regression, Kaplan–Meier survival and ROC analyses. Immune correlates were assessed in terms of immune cell infiltration, checkpoint and immunomodulatory activity. Drug sensitivity was predicted through molecular docking and molecular dynamics simulations to evaluate binding affinity and complex stability. Experimental validation was conducted in HNSCC cell lines.</div></div><div><h3>Results</h3><div>ACOX3 was significantly upregulated in KICH, PRAD and THCA, and downregulated in COAD, HNSCC, KIRP, LIHC and STAD. The OS Cox regression showed high ACOX3 expression was associated with a favorable prognosis in HNSCC but poor outcomes in LGG and UVM. The ROC curves showed that the AUC for ESCA, GBM, OV, PAAD, STES and WT exceeded 0.8. ACOX3 expression positively correlated with CD4⁺T, CD8⁺T and NK cells in HNSCC. Single-cell and spatial transcriptomics revealed ACOX3 enrichment in malignant regions, particularly in CD4⁺T, CD8⁺T and CD8⁺Tex cells. Drug screening prioritized AZD6482 and TGX-221 as high-affinity ACOX3 inhibitors, and the AZD6482 showed stable binding in MD simulations. Functional experiments confirmed that ACOX3 overexpression suppressed HNSCC cell proliferation, invasion and migration.</div></div><div><h3>Conclusion</h3><div>ACOX3 represents a dual diagnostic and prognostic biomarker with broad pan-cancer relevance, exhibiting distinct immune correlates and therapeutic potential.</div></div>","PeriodicalId":10616,"journal":{"name":"Computational Biology and Chemistry","volume":"121 ","pages":"Article 108844"},"PeriodicalIF":3.1,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145783855","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-04-01Epub Date: 2025-12-16DOI: 10.1016/j.compbiolchem.2025.108848
Tapas Ranjan Samala , Kunal Santosh Patil , Ethan Thomas John , Ramesh Eerapagula , Ajay Kumar Mahato , Priyankar Sen
Diabetes mellitus is a chronic metabolic disorder characterized by elevated blood glucose levels, and it poses significant health challenges globally. An established way of managing diabetes is through the use of α-amylase inhibitors. This study aimed to identify novel α-amylase inhibitors from phytochemicals identified from Halodule uninervis rhizomes for the control of postprandial blood glucose levels in individuals with type 2 diabetes. HRLCMS- and GCMS- identified analytes were screened on the basis of their ADME properties. Further screening was carried out on the basis of site-specific molecular docking with the α-amylase target and ligand combinations. The top 5 ranked dockings of each of the targets were further subjected to molecular dynamics simulations and analysis. On the basis of screening and molecular dynamics simulations, a glycosyloxyisoflavone, 6”-O-acetyldaidizen (6OAD) and an anthraquinone, frangulin B were found to be potential inhibitors of α-amylase on the basis of their interactions with the catalytic triad: ASP-197, GLU-233 and ASP-300. They interact via stable hydrogen bonding interactions with these residues at the enzymatic cleavage site of glycosylation. These findings suggest that 6”-O-acetyldaidizen (6OAD) and frangulin B possess both structural and dynamic attributes that are favourable for their use as putative type II diabetes therapeutics, via the regulation of postprandial glucose levels
糖尿病是一种以血糖水平升高为特征的慢性代谢性疾病，在全球范围内构成了重大的健康挑战。α-淀粉酶抑制剂是治疗糖尿病的一种有效方法。本研究旨在从盐菜根茎中鉴定出的植物化学物质中鉴定出新的α-淀粉酶抑制剂，用于控制2型糖尿病患者餐后血糖水平。根据其ADME性质对HRLCMS和GCMS鉴定的分析物进行筛选。在与α-淀粉酶靶点和配体组合进行位点特异性分子对接的基础上进行进一步筛选。对每个靶点的前5位进行分子动力学模拟和分析。通过筛选和分子动力学模拟，发现糖基氧异黄酮6′- o -乙酰基daidizen （6OAD）和蒽醌frangulin B与催化三元体ASP-197、GLU-233和ASP-300相互作用，是α-淀粉酶的潜在抑制剂。它们通过稳定的氢键作用与糖基化酶裂解位点的残基相互作用。这些发现表明，6 ' - o -乙酰代二酮（6OAD）和frangulin B具有结构和动态特性，通过调节餐后血糖水平，有利于它们作为2型糖尿病的治疗药物
{"title":"6-O-acetyldaidzen and frangulin B from Halodule uninervis as novel α-amylase inhibitors: A molecular dynamics perspective","authors":"Tapas Ranjan Samala , Kunal Santosh Patil , Ethan Thomas John , Ramesh Eerapagula , Ajay Kumar Mahato , Priyankar Sen","doi":"10.1016/j.compbiolchem.2025.108848","DOIUrl":"10.1016/j.compbiolchem.2025.108848","url":null,"abstract":"<div><div>Diabetes mellitus is a chronic metabolic disorder characterized by elevated blood glucose levels, and it poses significant health challenges globally. An established way of managing diabetes is through the use of α-amylase inhibitors. This study aimed to identify novel α-amylase inhibitors from phytochemicals identified from <em>Halodule uninervis</em> rhizomes for the control of postprandial blood glucose levels in individuals with type 2 diabetes. HR<img>LCMS- and GC<img>MS- identified analytes were screened on the basis of their ADME properties. Further screening was carried out on the basis of site-specific molecular docking with the α-amylase target and ligand combinations. The top 5 ranked dockings of each of the targets were further subjected to molecular dynamics simulations and analysis. On the basis of screening and molecular dynamics simulations, a glycosyloxyisoflavone, 6”-O-acetyldaidizen (6OAD) and an anthraquinone, frangulin B were found to be potential inhibitors of α-amylase on the basis of their interactions with the catalytic triad: ASP-197, GLU-233 and ASP-300. They interact via stable hydrogen bonding interactions with these residues at the enzymatic cleavage site of glycosylation. These findings suggest that 6”-O-acetyldaidizen (6OAD) and frangulin B possess both structural and dynamic attributes that are favourable for their use as putative type II diabetes therapeutics, via the regulation of postprandial glucose levels</div></div>","PeriodicalId":10616,"journal":{"name":"Computational Biology and Chemistry","volume":"121 ","pages":"Article 108848"},"PeriodicalIF":3.1,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145787165","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-04-01Epub Date: 2025-12-27DOI: 10.1016/j.compbiolchem.2025.108868
Chandramohan Nithya , Neelesh Babu Thummadi , P. Manimaran
Cancer remains a major global health challenge, underscoring the need to identify novel and effective therapeutic targets. In this study, we constructed a high-confidence cancer protein–protein interaction network and selected the largest connected component, comprising 2564 cancer-associated proteins linked by 20,747 interactions. We then evaluated 11 centrality measures to quantify the node importance. Using the TOPSIS multi-criteria decision-making approach, we ranked 2564 cancer-associated genes and identified the top 1 % (26 genes) as high-priority candidates. Drug–target mapping showed that 21 of these genes were associated with approved, investigational, or experimental drugs, whereas five genes, namely NXF1, CDC5L, MOV10, EP300, and CUL7 had no known therapeutic associations, marking them as unexplored targets. GO and KEGG enrichment analyses indicated roles in transcriptional regulation, RNA processing, ubiquitin-mediated protein degradation, and pathways such as Notch, JAK-STAT, and mRNA surveillance. The perturbations in these themes are increasingly associated with cancer development and progression, highlighting the possible roles of these genes in cancers. Survival analysis across multiple cancer types using TCGA datasets revealed significant prognostic effects: CDC5L was associated with improved survival in acute myeloid leukemia (hazard ratio (HR) = 0.59), EP300 expression correlated with better outcomes in kidney renal clear cell carcinoma (HR = 0.52), and elevated MOV10 expression predicted poor prognosis in kidney renal clear cell carcinoma (HR=2.5), lung adenocarcinoma (HR=1.5), and liver hepatocellular carcinoma (HR=1.5). Overexpression of CUL7 correlated with poor prognosis in colon adenocarcinoma (HR=2), and glioblastoma (HR=1.6). NXF1 showed cancer-type-specific results, associated with better prognosis in cervical cancer (HR=0.53) but poor prognosis in kidney renal clear cell carcinoma (HR=1.4). These findings provide quantitative evidence supporting the biological and clinical relevance of the prioritized genes, and the five untargeted genes emerge as strong candidates for future experimental validation through CRISPR-based perturbation, gene silencing, and functional phenotypic assays. Overall, this integrative TOPSIS-network framework offers a robust and reproducible strategy for uncovering both established and novel therapeutic targets, expanding the landscape for precision oncology.
{"title":"Network centrality–driven TOPSIS approach for prioritizing cancer therapeutic targets","authors":"Chandramohan Nithya , Neelesh Babu Thummadi , P. Manimaran","doi":"10.1016/j.compbiolchem.2025.108868","DOIUrl":"10.1016/j.compbiolchem.2025.108868","url":null,"abstract":"<div><div>Cancer remains a major global health challenge, underscoring the need to identify novel and effective therapeutic targets. In this study, we constructed a high-confidence cancer protein–protein interaction network and selected the largest connected component, comprising 2564 cancer-associated proteins linked by 20,747 interactions. We then evaluated 11 centrality measures to quantify the node importance. Using the TOPSIS multi-criteria decision-making approach, we ranked 2564 cancer-associated genes and identified the top 1 % (26 genes) as high-priority candidates. Drug–target mapping showed that 21 of these genes were associated with approved, investigational, or experimental drugs, whereas five genes, namely NXF1, CDC5L, MOV10, EP300, and CUL7 had no known therapeutic associations, marking them as unexplored targets. GO and KEGG enrichment analyses indicated roles in transcriptional regulation, RNA processing, ubiquitin-mediated protein degradation, and pathways such as Notch, JAK-STAT, and mRNA surveillance. The perturbations in these themes are increasingly associated with cancer development and progression, highlighting the possible roles of these genes in cancers. Survival analysis across multiple cancer types using TCGA datasets revealed significant prognostic effects: CDC5L was associated with improved survival in acute myeloid leukemia (hazard ratio (HR) = 0.59), EP300 expression correlated with better outcomes in kidney renal clear cell carcinoma (HR = 0.52), and elevated MOV10 expression predicted poor prognosis in kidney renal clear cell carcinoma (HR=2.5), lung adenocarcinoma (HR=1.5), and liver hepatocellular carcinoma (HR=1.5). Overexpression of CUL7 correlated with poor prognosis in colon adenocarcinoma (HR=2), and glioblastoma (HR=1.6). NXF1 showed cancer-type-specific results, associated with better prognosis in cervical cancer (HR=0.53) but poor prognosis in kidney renal clear cell carcinoma (HR=1.4). These findings provide quantitative evidence supporting the biological and clinical relevance of the prioritized genes, and the five untargeted genes emerge as strong candidates for future experimental validation through CRISPR-based perturbation, gene silencing, and functional phenotypic assays. Overall, this integrative TOPSIS-network framework offers a robust and reproducible strategy for uncovering both established and novel therapeutic targets, expanding the landscape for precision oncology.</div></div>","PeriodicalId":10616,"journal":{"name":"Computational Biology and Chemistry","volume":"121 ","pages":"Article 108868"},"PeriodicalIF":3.1,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145879619","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-04-01Epub Date: 2025-11-26DOI: 10.1016/j.compbiolchem.2025.108803
S. Sindhu, Shubashini K. Sripathi
Search for plant-based remedies for hypoestrogenism has validated the role of phytoestrogens as effective alternatives to endogenous estrogens. Extracts of Butea monosperma and its isolated metabolites have shown gynaecological effects pertaining to contraception, antifertility, anti-estrogenic and estrogenic potential as assessed by in vivo and in vitro assays. However, the molecular mechanism of interactions of these metabolites with estrogen receptors remains unexplored. To bridge this knowledge gap, the current research sought to characterize the pharmacological profile of flavonoids of this plant by an integrated computational approach. Twenty-nine compounds elaborated by the medicinal plant Butea monosperma were analysed for energy optimization using Guassian 16 software and taken up for molecular docking analysis by Schrodinger Maestro software and the binding affinity between the ligand and estrogenic receptors ER α - 1A52 and ER β - 3OLS were analysed. Molecular dynamic trajectories and prime mmGBSA binding free energy calculations were evaluated for ligands selected from the binding interaction score. Qikprop software and the protox server predicted ADME characteristics and toxicity of the molecules respectively. The molecular docking analysis demonstrated that six compounds displayed docking affinities comparable to that of the endogenous ligand 17β-estradiol at ER α, whereas fifteen compounds exhibited similar binding affinities at ER β. Furthermore, five compounds exhibited stronger binding affinities than 17β-estradiol toward ER α, while another five demonstrated enhanced binding affinities toward ER β, suggesting their potential as more efficacious receptor ligands. The compound catechin and isocoreopsin exhibited glide energy of −40.035 kcal/mol and −49.11 kal/mol at ER α respectively whereas 17β estradiol exhibited −52.012 kcal/mol. At ERβ, catechin and butin exhibited appreciable glide energy. These compounds were found to interact with amino acid residues HIS524, GLU353, PHE404, PHE356 and ARG346 similar to that of 17β estradiol. The study also revealed that chalcones and flavonols of Butea monosperma exhibit higher binding affinity to estrogenic receptors than the soy isoflavones genistein and daidzein.
{"title":"In silico evaluation of the estrogenic activity of flavonoids from Butea monosperma: Exploring phytoestrogenic alternatives to endogenous estrogens","authors":"S. Sindhu, Shubashini K. Sripathi","doi":"10.1016/j.compbiolchem.2025.108803","DOIUrl":"10.1016/j.compbiolchem.2025.108803","url":null,"abstract":"<div><div>Search for plant-based remedies for hypoestrogenism has validated the role of phytoestrogens as effective alternatives to endogenous estrogens. Extracts of <em>Butea monosperma</em> and its isolated metabolites have shown gynaecological effects pertaining to contraception, antifertility, anti-estrogenic and estrogenic potential as assessed by <em>in vivo</em> and <em>in vitro</em> assays. However, the molecular mechanism of interactions of these metabolites with estrogen receptors remains unexplored. To bridge this knowledge gap, the current research sought to characterize the pharmacological profile of flavonoids of this plant by an integrated computational approach. Twenty-nine compounds elaborated by the medicinal plant <em>Butea monosperma</em> were analysed for energy optimization using Guassian 16 software and taken up for molecular docking analysis by Schrodinger Maestro software and the binding affinity between the ligand and estrogenic receptors ER α - 1A52 and ER β - 3OLS were analysed. Molecular dynamic trajectories and prime mmGBSA binding free energy calculations were evaluated for ligands selected from the binding interaction score. Qikprop software and the protox server predicted ADME characteristics and toxicity of the molecules respectively. The molecular docking analysis demonstrated that six compounds displayed docking affinities comparable to that of the endogenous ligand 17β-estradiol at ER α, whereas fifteen compounds exhibited similar binding affinities at ER β. Furthermore, five compounds exhibited stronger binding affinities than 17β-estradiol toward ER α, while another five demonstrated enhanced binding affinities toward ER β, suggesting their potential as more efficacious receptor ligands. The compound catechin and isocoreopsin exhibited glide energy of −40.035 kcal/mol and −49.11 kal/mol at ER α respectively whereas 17β estradiol exhibited −52.012 kcal/mol. At ERβ, catechin and butin exhibited appreciable glide energy. These compounds were found to interact with amino acid residues HIS524, GLU353, PHE404, PHE356 and ARG346 similar to that of 17β estradiol. The study also revealed that chalcones and flavonols of <em>Butea monosperma</em> exhibit higher binding affinity to estrogenic receptors than the soy isoflavones genistein and daidzein.</div></div>","PeriodicalId":10616,"journal":{"name":"Computational Biology and Chemistry","volume":"121 ","pages":"Article 108803"},"PeriodicalIF":3.1,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145727713","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-04-01Epub Date: 2025-11-24DOI: 10.1016/j.compbiolchem.2025.108801
Alexandre Luiz Korte de Azevedo, Mateus Vinicius Oliveira Pereira, Enilze Maria de Souza Fonseca Ribeiro, Talita Helen Bombardelli Gomig
Breast cancer heterogeneity stems from diverse molecular alterations, including proteostasis loss due to chaperone system dysfunction. However, the impact of impaired chaperone activity on proteomic changes and tumorigenesis remains unclear. Here, characterized the expression patterns of major chaperone families and mapped their client protein interactions to elucidate their role in shaping tumor biology. We identified 53 chaperones expressed in breast tissue, of which 26 were differentially expressed between tumor and non-tumor samples. Using validated protein interaction data and machine-learning predictions, coupled with molecular docking, we constructed protein-protein interaction (PPI) networks for each chaperone family and subsequently performed enrichment analyses to assess their involvement in cancer-related pathways. Each chaperone family’s PPI network comprised a distinct set of client proteins and was enriched in different biological pathways and processes. The HSP70 system PPI network included LYN, NFKB1, and PARP1, and was related to DNA repair and immunomodulation through interleukin and cytokine signaling. Although a partial overlap of client proteins was observed between the HSP70 and HSP90 sets, HSP90 was also associated with particular client proteins, including TRAF2, PDGFRB, and NUDC, which were enriched in MAPK and PI3K/AKT/mTOR signaling pathways, as well as epithelial-to-mesenchymal transition and cell cycle control. Our results also indicate an association between CCT/TRiC chaperonins and the regulation of tubulin/actin, supporting their involvement in cytoskeleton dynamics, the mitotic spindle, chromosome segregation, and autophagy/aggrephagy. Overall, our findings expand the repertoire of chaperone client proteins and provide insights into how chaperone dysregulation influence breast cancer biology, highlighting their potential as therapeutic targets.
{"title":"Functional profiling of the chaperone systems interactome in breast cancer using experimental and machine-learning data","authors":"Alexandre Luiz Korte de Azevedo, Mateus Vinicius Oliveira Pereira, Enilze Maria de Souza Fonseca Ribeiro, Talita Helen Bombardelli Gomig","doi":"10.1016/j.compbiolchem.2025.108801","DOIUrl":"10.1016/j.compbiolchem.2025.108801","url":null,"abstract":"<div><div>Breast cancer heterogeneity stems from diverse molecular alterations, including proteostasis loss due to chaperone system dysfunction. However, the impact of impaired chaperone activity on proteomic changes and tumorigenesis remains unclear. Here, characterized the expression patterns of major chaperone families and mapped their client protein interactions to elucidate their role in shaping tumor biology. We identified 53 chaperones expressed in breast tissue, of which 26 were differentially expressed between tumor and non-tumor samples. Using validated protein interaction data and machine-learning predictions, coupled with molecular docking, we constructed protein-protein interaction (PPI) networks for each chaperone family and subsequently performed enrichment analyses to assess their involvement in cancer-related pathways. Each chaperone family’s PPI network comprised a distinct set of client proteins and was enriched in different biological pathways and processes. The HSP70 system PPI network included LYN, NFKB1, and PARP1, and was related to DNA repair and immunomodulation through interleukin and cytokine signaling. Although a partial overlap of client proteins was observed between the HSP70 and HSP90 sets, HSP90 was also associated with particular client proteins, including TRAF2, PDGFRB, and NUDC, which were enriched in MAPK and PI3K/AKT/mTOR signaling pathways, as well as epithelial-to-mesenchymal transition and cell cycle control. Our results also indicate an association between CCT/TRiC chaperonins and the regulation of tubulin/actin, supporting their involvement in cytoskeleton dynamics, the mitotic spindle, chromosome segregation, and autophagy/aggrephagy. Overall, our findings expand the repertoire of chaperone client proteins and provide insights into how chaperone dysregulation influence breast cancer biology, highlighting their potential as therapeutic targets.</div></div>","PeriodicalId":10616,"journal":{"name":"Computational Biology and Chemistry","volume":"121 ","pages":"Article 108801"},"PeriodicalIF":3.1,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145600463","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-04-01Epub Date: 2025-12-19DOI: 10.1016/j.compbiolchem.2025.108864
Yang Gao , Haoyun Yu , Chunman Zuo
Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry (MALDI-MS) is a powerful tool for profiling complex biological samples. However, large-scale metabolomics experiments often suffer from substantial batch effects caused by variations in sample processing, instrument conditions, and acquisition protocols. These non-biological variations obscure true biological signals, reduce reproducibility, and compromise the generalizability of downstream models. Existing correction methods either rely on oversimplified linear assumptions or risk over-correcting and removing meaningful biological differences. To address this challenge, we propose HGAlign(Heterogeneous Graph Alignment Model), a neural network model that corrects batch effects in large-scale MALDI-MS experiments while preserving important biological differences. Our approach uses heterogeneous graph convolutional networks to learn relationships between samples and metabolic features, enabling effective batch correction without losing disease-related information.
Extensive experiments on CyTOF public datasets and clinical MALDI-MS serum data from systemic lupus erythematosus (SLE) patients demonstrate that HGAlign significantly reduces inter-batch discrepancies while maintaining or improving classification accuracy. Quantitative evaluation shows that our method achieves the lowest MMD values among state-of-the-art methods, and consistently improves classification metrics. Moreover, HGAlign avoids over-correction, enabling stable identification of cross-batch differential metabolites that retain biological interpretability.
HGAlign offers a principled framework for balancing batch effect removal and biological signal preservation in high-throughput metabolomics. By introducing heterogeneous graph representation learning, it achieves superior performance in both batch correction and disease classification tasks, showing strong potential for large-scale clinical applications.
{"title":"HGAlign: Biologically preserving batch correction and classification for metabolomics via heterogeneous graph alignment","authors":"Yang Gao , Haoyun Yu , Chunman Zuo","doi":"10.1016/j.compbiolchem.2025.108864","DOIUrl":"10.1016/j.compbiolchem.2025.108864","url":null,"abstract":"<div><div>Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry (MALDI-MS) is a powerful tool for profiling complex biological samples. However, large-scale metabolomics experiments often suffer from substantial batch effects caused by variations in sample processing, instrument conditions, and acquisition protocols. These non-biological variations obscure true biological signals, reduce reproducibility, and compromise the generalizability of downstream models. Existing correction methods either rely on oversimplified linear assumptions or risk over-correcting and removing meaningful biological differences. To address this challenge, we propose HGAlign(Heterogeneous Graph Alignment Model), a neural network model that corrects batch effects in large-scale MALDI-MS experiments while preserving important biological differences. Our approach uses heterogeneous graph convolutional networks to learn relationships between samples and metabolic features, enabling effective batch correction without losing disease-related information.</div><div>Extensive experiments on CyTOF public datasets and clinical MALDI-MS serum data from systemic lupus erythematosus (SLE) patients demonstrate that HGAlign significantly reduces inter-batch discrepancies while maintaining or improving classification accuracy. Quantitative evaluation shows that our method achieves the lowest MMD values among state-of-the-art methods, and consistently improves classification metrics. Moreover, HGAlign avoids over-correction, enabling stable identification of cross-batch differential metabolites that retain biological interpretability.</div><div>HGAlign offers a principled framework for balancing batch effect removal and biological signal preservation in high-throughput metabolomics. By introducing heterogeneous graph representation learning, it achieves superior performance in both batch correction and disease classification tasks, showing strong potential for large-scale clinical applications.</div></div>","PeriodicalId":10616,"journal":{"name":"Computational Biology and Chemistry","volume":"121 ","pages":"Article 108864"},"PeriodicalIF":3.1,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145835531","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-04-01Epub Date: 2025-11-27DOI: 10.1016/j.compbiolchem.2025.108804
Yang Yuan , Jianqiang Du , Yanchen Zhu , Jigen Luo , Qiang Huang
In the field of metabolomics data analysis, missing values are a common challenge. Traditional k-nearest neighbors (KNN) imputation methods often overlook the distribution of the original data, resulting in suboptimal outcomes when addressing missing values in metabolomics. To better restore the data distribution and enhance the imputation results, this paper introduces a dynamic weighted KNN imputation algorithm with preprocessing based on the normal distribution (NDW-KNN). Initially, the similarity distance between samples is calculated to assign an appropriate k value to each sample. Subsequently, missing values are categorized based on the similarity of the neighbors of the target sample and undergo normal distribution preprocessing. Finally, an inverse distance weighting method is used to assign weights to each sample, thereby predicting missing values. Experimental results show that NDW-KNN achieved the best performance across three benchmark metabolomics datasets, reducing the average NRMSE and MAPE by 21.7 % and 32.9 % compared with traditional KNN, and by 4.5 % and 13.8 % compared with NS-KNN. Even under a missing rate as high as 30 %, NDW-KNN maintained the lowest imputation error and the highest consistency with the original data distribution, while exhibiting stronger intergroup discrimination in principal component analysis, demonstrating its excellent robustness and practical applicability.
{"title":"Research on the application of dynamic weighted KNN with preprocessing based on a normal distribution in metabolomics data imputation","authors":"Yang Yuan , Jianqiang Du , Yanchen Zhu , Jigen Luo , Qiang Huang","doi":"10.1016/j.compbiolchem.2025.108804","DOIUrl":"10.1016/j.compbiolchem.2025.108804","url":null,"abstract":"<div><div>In the field of metabolomics data analysis, missing values are a common challenge. Traditional k-nearest neighbors (KNN) imputation methods often overlook the distribution of the original data, resulting in suboptimal outcomes when addressing missing values in metabolomics. To better restore the data distribution and enhance the imputation results, this paper introduces a dynamic weighted KNN imputation algorithm with preprocessing based on the normal distribution (NDW-KNN). Initially, the similarity distance between samples is calculated to assign an appropriate k value to each sample. Subsequently, missing values are categorized based on the similarity of the neighbors of the target sample and undergo normal distribution preprocessing. Finally, an inverse distance weighting method is used to assign weights to each sample, thereby predicting missing values. Experimental results show that NDW-KNN achieved the best performance across three benchmark metabolomics datasets, reducing the average NRMSE and MAPE by 21.7 % and 32.9 % compared with traditional KNN, and by 4.5 % and 13.8 % compared with NS-KNN. Even under a missing rate as high as 30 %, NDW-KNN maintained the lowest imputation error and the highest consistency with the original data distribution, while exhibiting stronger intergroup discrimination in principal component analysis, demonstrating its excellent robustness and practical applicability.</div></div>","PeriodicalId":10616,"journal":{"name":"Computational Biology and Chemistry","volume":"121 ","pages":"Article 108804"},"PeriodicalIF":3.1,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145682574","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-04-01Epub Date: 2025-12-11DOI: 10.1016/j.compbiolchem.2025.108828
Rabia Kalkan Cakmak , Nail Besli , Nilufer Ercin , Ulkan Celik
Glucagon-like peptide-1 (GLP-1), a pivotal incretin hormone modulating glycemic homeostasis, has emerged as a clinically validated target for the treatment of type 2 diabetes and obesity. In this study, we present a comprehensive AI-integrated drug discovery pipeline that leverages BioBERT-based biomedical text mining to delineate the therapeutic landscape of GLP-1 receptor agonism systematically. Subsequent high-throughput virtual screening (HTVS) of a curated natural product library identified structurally diverse candidate ligands. A machine-learning-guided ADMET profiling algorithm was employed to prioritize compounds with optimal pharmacokinetic and safety characteristics. Top-ranked molecules were subjected to extensive molecular dynamics (MD) simulations using the GROMACS platform, enabling quantitative evaluation of structural stability, dynamic behavior, and receptor-ligand interaction persistence. Molecular docking analyses demonstrated robust binding affinities (ΔG: −11.3 to −8.7 kcal/mol), while MM-PBSA free energy estimations (ΔG<−30 kcal/mol) corroborated the thermodynamic favorability of binding. Among the screened entities, five lead candidates—CNP0244222.1, CNP0186692.11, CNP0361941.2, CNP0547477.1, and CNP0258197.2—consistently exhibited superior ADMET scores (>0.67), stable interaction trajectories, and enthalpically favorable profiles. This integrative, AI-augmented computational framework demonstrates substantial potential to accelerate the rational design and preclinical advancement of GLP-1-targeted therapeutics.
{"title":"AI-powered literature mining reveals the therapeutic significance of GLP-1 receptor: Simulation of natural agonist candidates based on molecular dynamics","authors":"Rabia Kalkan Cakmak , Nail Besli , Nilufer Ercin , Ulkan Celik","doi":"10.1016/j.compbiolchem.2025.108828","DOIUrl":"10.1016/j.compbiolchem.2025.108828","url":null,"abstract":"<div><div>Glucagon-like peptide-1 (GLP-1), a pivotal incretin hormone modulating glycemic homeostasis, has emerged as a clinically validated target for the treatment of type 2 diabetes and obesity. In this study, we present a comprehensive AI-integrated drug discovery pipeline that leverages BioBERT-based biomedical text mining to delineate the therapeutic landscape of GLP-1 receptor agonism systematically. Subsequent high-throughput virtual screening (HTVS) of a curated natural product library identified structurally diverse candidate ligands. A machine-learning-guided ADMET profiling algorithm was employed to prioritize compounds with optimal pharmacokinetic and safety characteristics. Top-ranked molecules were subjected to extensive molecular dynamics (MD) simulations using the GROMACS platform, enabling quantitative evaluation of structural stability, dynamic behavior, and receptor-ligand interaction persistence. Molecular docking analyses demonstrated robust binding affinities (ΔG: −11.3 to −8.7 kcal/mol), while MM-PBSA free energy estimations (ΔG<−30 kcal/mol) corroborated the thermodynamic favorability of binding. Among the screened entities, five lead candidates—CNP0244222.1, CNP0186692.11, CNP0361941.2, CNP0547477.1, and CNP0258197.2—consistently exhibited superior ADMET scores (>0.67), stable interaction trajectories, and enthalpically favorable profiles. This integrative, AI-augmented computational framework demonstrates substantial potential to accelerate the rational design and preclinical advancement of GLP-1-targeted therapeutics.</div></div>","PeriodicalId":10616,"journal":{"name":"Computational Biology and Chemistry","volume":"121 ","pages":"Article 108828"},"PeriodicalIF":3.1,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145734065","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The study investigates 5- ((2-nitrobenzylidene) amino 2-phenyloxazole-4-carbonitrile (PS13), a derivative of the oxazole that was designed to block the ERBB3 receptor that plays a role in breast cancer development. The syntheses of PS13 were performed in two steps due to condensation and its structure was verified with the help of IR NMR, MS, and elemental analysis. Strong binding affinity was observed between the molecules and ERBB3 with the docking score of −9.5 kcal/mol that was reinforced by the presence of key hydrogen and hydrophobic bonds. Simulation of molecular dynamics above 500 ns showed that the formation of the ligand-receptor complex was stable, and the fluctuations of RMSD were minimal, which proves the structural compatibility of the molecules and the stability of their interaction. The ADMET profiling predicted good drug-like, gastrointestinal absorption, non-P-gp substrate, and good metabolism. The analysis of density functional theory indicated that the HOMO-LUMO energy gap is −2.27 eV, which indicated the stability of the electronics, and the ability to be reactive. The PS13-SLNs that were developed were PS13-loaded solid lipid nanoparticles that had high encapsulation efficiency (81 +/- 2.16 %), and enhanced release profiles in both the acidic and neutral pH conditions. Both in vitro MTT assays of MCF-7 cells and morphological changes depicted the dose-dependent cytotoxicity with 60.27 ± 0.04 µg/mL of IC50, and morphological changes that were consonant to apoptosis. Drug release kinetics indicated a first-order mechanism and Fickian diffusion, suggesting a controlled release profile. All these combined with the high ERBB3 binding affinity, good pharmacokinetics, stable SLN formulation, and in vitro anticancer efficacy of PS13, indicate that PS13 is a promising lead candidate to advance in preclinical development in the treatment of breast cancer.
{"title":"Targeted anticancer potential of oxazole derivative against breast cancer: Synthesis, molecular docking, dynamics simulation, and in vitro evaluation on ERBB3 receptor","authors":"Jianxing Xu , Dongwei Zhu , Kanagaraj Rajalakshmi , Mangirish Deshpande , Natarajan Kiruthiga , Panneerselvam Theivendren , Selvaraj Muthusamy , Siyi Wu , Weizhong Zhao","doi":"10.1016/j.compbiolchem.2025.108859","DOIUrl":"10.1016/j.compbiolchem.2025.108859","url":null,"abstract":"<div><div>The study investigates 5- ((2-nitrobenzylidene) amino 2-phenyloxazole-4-carbonitrile (PS13), a derivative of the oxazole that was designed to block the ERBB3 receptor that plays a role in breast cancer development. The syntheses of PS13 were performed in two steps due to condensation and its structure was verified with the help of IR NMR, MS, and elemental analysis. Strong binding affinity was observed between the molecules and ERBB3 with the docking score of −9.5 kcal/mol that was reinforced by the presence of key hydrogen and hydrophobic bonds. Simulation of molecular dynamics above 500 ns showed that the formation of the ligand-receptor complex was stable, and the fluctuations of RMSD were minimal, which proves the structural compatibility of the molecules and the stability of their interaction. The ADMET profiling predicted good drug-like, gastrointestinal absorption, non-P-gp substrate, and good metabolism. The analysis of density functional theory indicated that the HOMO-LUMO energy gap is −2.27 eV, which indicated the stability of the electronics, and the ability to be reactive. The PS13-SLNs that were developed were PS13-loaded solid lipid nanoparticles that had high encapsulation efficiency (81 +/- 2.16 %), and enhanced release profiles in both the acidic and neutral pH conditions. Both in vitro MTT assays of MCF-7 cells and morphological changes depicted the dose-dependent cytotoxicity with 60.27 ± 0.04 µg/mL of IC<sub>50</sub>, and morphological changes that were consonant to apoptosis. Drug release kinetics indicated a first-order mechanism and Fickian diffusion, suggesting a controlled release profile. All these combined with the high ERBB3 binding affinity, good pharmacokinetics, stable SLN formulation, and in vitro anticancer efficacy of PS13, indicate that PS13 is a promising lead candidate to advance in preclinical development in the treatment of breast cancer.</div></div>","PeriodicalId":10616,"journal":{"name":"Computational Biology and Chemistry","volume":"121 ","pages":"Article 108859"},"PeriodicalIF":3.1,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145866930","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号