Bioinformatics and Biology Insights最新文献

{"title":"Integrative Analysis of Intestinal Transcriptomes Underscores <i>C2</i> Upregulation in IBD, AIG, EoE, and CRC.","authors":"Arman Mokaram Doust Delkhah","doi":"10.1177/11779322261476630","DOIUrl":"10.1177/11779322261476630","url":null,"abstract":"<p><p>Intestinal transcriptomic data encompass substantial value in studying inflammatory bowel disease. By analyzing intestinal gene expression profiles, this study aimed to discover transcripts exhibiting diagnostic efficacy for IBD. Intestinal transcriptomic data from 1,458 patients with Crohn's disease (CD), 1,211 patients with ulcerative colitis (UC), and 674 healthy controls were acquired by integrating seven datasets with accession numbers GSE66407, GSE193677, GSE126124, GSE83687, GSE75214, GSE36807, and GSE16879. Then, the DEGs in IBD were analyzed by ML methods to discern transcripts that hold diagnostic power. ROC analysis identified potential biomarkers, which were subsequently tested in 16 external datasets. <i>C2</i>, <i>NLRC5</i>, <i>S100P</i>, <i>PGAP3</i>, and <i>GPR15</i> emerged as prominent genes based on both RF and LASSO methods. Meanwhile, only <i>NLRC5</i> and <i>C2</i> had an AUC of the ROC curve greater than the 0.7 threshold in the integrated data. ROC analysis across 23 cohorts demonstrated that the AUC of <i>C2</i> and <i>NLRC5</i> were above 0.7 in the majority of the cohorts, particularly that for <i>C2</i>. However, upregulated <i>C2</i> also exhibited diagnostic efficacy for autoimmune gastritis (AIG), eosinophilic esophagitis (EoE), and colorectal cancer (CRC), reflecting the substantial activation of the complement cascade in these disorders manifesting with gastrointestinal inflammation.</p>","PeriodicalId":9065,"journal":{"name":"Bioinformatics and Biology Insights","volume":"20 ","pages":"11779322261476630"},"PeriodicalIF":2.0,"publicationDate":"2026-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13438343/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148677093","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Integrated in Silico and in Vitro Bioactivity Profiling of Selected South African Plant Constituents Against Carbohydrate-Metabolizing Enzymes for Type 2 Diabetes Therapy.","authors":"Oluwaseye Adedirin, Saheed Sabiu","doi":"10.1177/11779322261455977","DOIUrl":"10.1177/11779322261455977","url":null,"abstract":"<p><p>The increasing global burden of Type 2 diabetes mellitus (T2DM) and the side effects of conventional drugs have driven the search for natural therapeutic alternatives. This study investigates the antidiabetic potential of South African essential oils - specifically <i>Artemisia afra, Agathosma betulina, and Cymbopogon citratus</i> - which are traditionally used in diabetes treatment but lack comprehensive scientific validation. Using an integrated research framework: mining the phytochemical metabolites of the plant source from Dr. Duke's Phytochemical and Ethnobotanical Database and relevant literature; screening of metabolites using molecular docking and molecular dynamics (MD) to identify potential leads; and <i>in vitro</i> enzymatic assays of the oils for α-amylase (AA) and α-glucosidase (AG) inhibition. Computational results identified quercetin-3,7-diglucoside (Q37DG), quercetin-7-O-glucoside (Q7G), and rutin as promising lead compounds. Notably, Q37DG demonstrated favorable inhibitory potential, with ΔG_bind values of -52.73 kcal/mol for AA and -36.60 kcal/mol for AG, exceeding the known inhibitor acarbose (-43.59 and -34.41 kcal/mol, respectively). MD simulations further supported Q37DG's role in enhancing enzyme stability upon binding, indicating a competitive inhibition mechanism through strong interactions with catalytic residues such as ASP469 (AG) and ASP197 (AA). Genetic Algorithm-based Quantitative Structure-Activity Relationship (GA-QSAR) models predict impressive IC₅₀ values for Q37DG (3.47 μM for AG and 0.68 μM for AA) compared to acarbose (6.92 and 2.51 μM). Experimental <i>in vitro</i> assays confirmed <i>A. betulina</i> essential oil as the oil with the most activity (IC₅₀ = 6.02 μg/ml for AA and 6.70 μg/ml for AG). Surprisingly, the lead compounds identified via computational screening are reported constituents of <i>A. betulina</i>, thereby validating the integrated data mining and in silico approach. This work promotes green chemistry by valuing indigenous plant resources and suggests that these leads warrant further investigation for nano-formulation and therapeutic development.</p>","PeriodicalId":9065,"journal":{"name":"Bioinformatics and Biology Insights","volume":"20 ","pages":"11779322261455977"},"PeriodicalIF":2.0,"publicationDate":"2026-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13438231/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148677090","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Unveiling Carbonic Anhydrase VIII, X, XI Expression in Cancer and Neurological Diseases Through Integrated Bioinformatics Approaches.","authors":"Ratul Bhowmik, Rajarshi Ray, Ashok Aspatwar","doi":"10.1177/11779322261475822","DOIUrl":"10.1177/11779322261475822","url":null,"abstract":"<p><p>Carbonic anhydrases are metalloenzymes found both in vertebrates and invertebrates. The CAs catalyze the reversible hydration of CO₂ to bicarbonate and H+ ions and play a significant role in respiration, transport of CO₂, pH homeostasis, electrolyte secretion, and biosynthetic reactions. The enzymatic activity of CAs is due to the coordination of Zn²⁺ in the active site by three histidine residues; however, in humans, there are three CAs known as CA-related proteins (CARPs) that are catalytically inactive due to the absence of one or more of the three histidine residues required for the coordination of the Zn²⁺ in the active site. Studies have shown that CARPs are expressed in all parts of the brain and are overexpressed in some cancers suggesting that the CARPs play a crucial role in neurological disorders and the development of cancer. However, the precise physiological roles of CARPs are still an enigma. In this study, we present a comprehensive biological workflow that employs various machine learning methodologies and statistical procedures to assess the similarity across CARPs by evaluating shared biological parameters. This approach enabled us to identify potential biomarkers, including transcription factors, co-expressed genes, and phenotypes, that may influence the expression of CARPs within disease pathways. Furthermore, we proposed a computational human health model by analyzing drugs and chemical candidates to prioritize compounds that may modulate regulatory networks associated with CARPs. These computational analyses identified candidate compounds for future experimental investigation in the context of neurological disorders and cancer.</p>","PeriodicalId":9065,"journal":{"name":"Bioinformatics and Biology Insights","volume":"20 ","pages":"11779322261475822"},"PeriodicalIF":2.0,"publicationDate":"2026-08-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13434096/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148673070","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Bioinformatics Analysis of the PSD Gene Family: Metastasis and Prognosis in HCC.","authors":"Hongyi Sun, Minghao Guo, Wei Liu","doi":"10.1177/11779322261425454","DOIUrl":"10.1177/11779322261425454","url":null,"abstract":"<p><p>Hepatocellular carcinoma (HCC) is a common malignant tumor with a poor prognosis for advanced metastasis while there is currently a lack of prognostic markers and effective therapeutic targets in this area. The <i>PSD</i> gene family, comprising four genes, has been implicated in tumor metastasis in breast cancer, but its role in HCC remains unclear. To elucidate the role-played by the <i>PSD</i> gene family in the prognosis of HCC and explore the relationship between the <i>PSD</i> gene family and tumor metastasis, we conducted a bioinformatics analysis using TCGA and gene expression omnibus (GEO) databases to investigate <i>PSD</i> gene family expression in HCC and its correlation with patient survival. Furthermore, we performed quantitative real-time polymerase chain reaction (qPCR) validation of the expression levels of the <i>PSD</i> gene family in clinical samples. Gene correlation and enrichment analyses were also performed to explore its relationship with tumor metastasis. In conclusion, the expression levels of the <i>PSD</i> gene family are significantly downregulated in HCC (<i>P</i> < 0.01), and the lower expressions are associated with poor survival. In addition, there is a link between the <i>PSD</i> gene family and RAS pathway activation, suggesting its potential role in metastasis.</p>","PeriodicalId":9065,"journal":{"name":"Bioinformatics and Biology Insights","volume":"20 ","pages":"11779322261425454"},"PeriodicalIF":2.0,"publicationDate":"2026-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13424774/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148652928","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Tracing the Evolutionary Dynamics and the Transcriptional Partitioning of Terpene Synthase Gene Family in <i>Coffea arabica</i> and Its Diploid Progenitors.","authors":"Makarim Elfadil M Osman, Mohanad A Ibrahim, Rieham Sallah H Osman, Budour A Alghamdi, Amina I Dirar, Wafa A Alrasheed, Abdelsattar M Omar, Emadeldin Hassan E Konozy","doi":"10.1177/11779322261472433","DOIUrl":"10.1177/11779322261472433","url":null,"abstract":"<p><p>Terpenes represent a diverse class of plant secondary metabolites that play important roles in development, ecological interactions, and aroma, and are synthesized by the terpene synthase (TPS) gene family. To investigate the evolutionary history and regulatory diversification of this family in coffee, we performed a genome-wide analysis of TPS genes in <i>Coffea arabica</i> and its diploid progenitors, <i>C. canephora</i> and <i>C. eugenioides</i>. Using domain-based searches targeting the PF01397 module, we identified 69 TPS genes in <i>C. arabica</i>, 40 in <i>C. canephora</i>, and 55 in <i>C. eugenioides</i>. Phylogenetic classification, duplication analysis, and synteny mapping revealed that the tandem duplication is the major driver of TPS expansion, with <i>C. arabica</i> retaining progenitor-derived copies alongside limited lineage-specific duplications. Selection analyses indicated that most TPS duplicates from <i>C. arabica</i> are under strong purifying constraint, while only a single orthogroup (OG0000013) showed limited evidence of episodic positive selection with a minor fraction of inferred sites and a lack of strong BEB support. Indicating that adaptive divergence was highly localized against a predominantly conserved evolutionary background. To link evolutionary divergence with transcriptional regulation, RNA-seq data from multiple tissues and developmental stages were analyzed. A substantial proportion of TPS genes were transcriptionally active and exhibited tissue- and stage-specific regulation. While many duplicates showed conserved or asymmetric expression, the OG0000013 duplicates combined localized adaptive sequence divergence with significant upregulation during late seed development. In conclusion, the retention and asymmetric transcription of progenitor-derived homeologs indicate that regulatory partitioning, rather than extensive gene-family innovation, has the dominant evolutionary outcome following allopolyploidization in <i>C. arabica</i>.</p>","PeriodicalId":9065,"journal":{"name":"Bioinformatics and Biology Insights","volume":"20 ","pages":"11779322261472433"},"PeriodicalIF":2.0,"publicationDate":"2026-07-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13400922/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148590372","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Hidden in the Pangenome? Machine Learning-Driven Discovery of Antimicrobial Potential in <i>Corynebacterium glutamicum</i>.","authors":"Sk Injamamul Islam, Khandker Shahed, Mst Mahmuda Parvin, Saloa Sanjida","doi":"10.1177/11779322261461933","DOIUrl":"10.1177/11779322261461933","url":null,"abstract":"<p><p>Antimicrobial resistance poses an increasing global challenge, driving the urgent need for alternative strategies to identify novel therapeutic agents. Microbial natural products encoded by biosynthetic gene clusters (BGCs) remain among the most promising sources of bioactive compounds. Although <i>Corynebacterium glutamicum</i> is best known as an industrial producer of amino acids, its potential as a producer of secondary metabolites has not been comprehensively assessed, despite the availability of numerous high-quality genome sequences. In this study, we carried out a comparative pangenome analysis of 36 complete <i>C. glutamicum</i> genomes and systematically mined for BGCs to explore the species' biosynthetic repertoire. Our analysis revealed variation in BGC content among strains, with several isolates harboring more hybrid clusters than others, suggesting metabolic diversity across the species. In addition to conserved terpene biosynthetic pathways, we detected polyketide-associated clusters not previously reported in <i>C. glutamicum</i>, expanding its recognized metabolic potential. RiPP-like clusters, including Lactococcin-related variants, were also identified, highlighting an underexplored reservoir of antimicrobials. To prioritize candidates for future validation, Support Vector Machine, Random Forest, and k-Nearest Neighbor models were trained on Composition, Transition, and Distribution (CTD) physicochemical sequence features and applied to genome-mined small open reading frames. The models demonstrated strong predictive performance, with the Support Vector Machine achieving the highest accuracy (84.1%), F1 score (83.8%), and area under the ROC curve (AUC = 0.920). After removing duplicate sequence IDs and applying a high-confidence AMP probability threshold (≥0.95), 18 unique AMP-like candidates were identified as promising. Overall, this study presents <i>C. glutamicum</i> as a promising source of bioactive metabolite candidates and shows how pangenome-scale mining combined with machine learning can support antimicrobial peptide discovery while still requiring experimental validation of the predicted leads.</p>","PeriodicalId":9065,"journal":{"name":"Bioinformatics and Biology Insights","volume":"20 ","pages":"11779322261461933"},"PeriodicalIF":2.0,"publicationDate":"2026-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13276223/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148293543","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"<i>In silico</i> Design and Analysis of Engineered Proteins Containing Multi-Epitope and Immunodominant Domains Derived From <i>Rickettsia prowazekii A</i>ntigens.","authors":"Massrehaw Assnakew Abebe, Desye Melese Sisay, Abebaw Fekadu, Abrham Tesfaye Bika, Alemu Tekewe Mogus, Tesfaye Sisay Tessema","doi":"10.1177/11779322261461934","DOIUrl":"10.1177/11779322261461934","url":null,"abstract":"<p><p>Epidemic louse-borne typhus caused by <i>Rickettsia prowazekii</i> and transmitted by the human body louse, remains a major public health threat in many developing regions. Historical records indicate that outbreaks have resulted in up to million cases annually, with mortality ranging from thousands to millions. Early and accurate diagnosis is critical for improving the efficacy of antibiotic therapies. However, existing serological diagnostic methods often suffer from limited sensitivity, specificity, and reliability. In this study, we applied integrated <i>in silico i</i>mmunoinformatics and structural bioinformatics approaches to identify antigenic targets and to design and characterize engineered proteins for diagnostic applications. Genomic and proteomic analyses identified Sca4 and OmpB as highly antigenic proteins of <i>Rickettsia prowazekii</i>. Computational epitope prediction mapped linear B-cell epitopes within multi-epitope regions of Sca4 and immunodominant regions of OmpB. Two engineered proteins (MA1 and MA2) were computationally designed: MA1 using selected linear B-cell epitopes linked via flexible peptide linkers with a fusion tag, and MA2 using truncated immunodominant domains with a fusion tag. The engineered proteins were computationally evaluated for physicochemical properties, antigenicity, solubility, stability, and sequence homology, and they demonstrated favorable characteristics with minimal similarity to human proteins. Structural modeling, followed by molecular docking and molecular dynamics simulations with the TLR4 receptor, suggested preserved structural integrity and stable protein-receptor interactions. Overall, these <i>in silico</i> findings underscore the potential of MA1 and MA2 constructs as potential diagnostic candidates and provide a basis for further experimental validation, including protein expression and purification, antibody production in animal models, and immunodiagnostic assay development.</p>","PeriodicalId":9065,"journal":{"name":"Bioinformatics and Biology Insights","volume":"20 ","pages":"11779322261461934"},"PeriodicalIF":2.0,"publicationDate":"2026-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13276219/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148293508","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Chemoinformatic Approaches to Identify Bioactive Inhibitors Against Type I Dehydroquinase (DHQ1) Enzyme of Typhoidal <i>Salmonella</i>.","authors":"Aqsa Ashfaq, Safah Mariyam, Ubair Aziz, Mamuna Mukhtar, Saadia Andleeb","doi":"10.1177/11779322261418889","DOIUrl":"10.1177/11779322261418889","url":null,"abstract":"<p><p>The incidence of antibiotic resistance is critically jeopardizing the public health of the new millennium, affecting both clinical and therapeutic outcomes. Typhoid perforation in low- and middle-income countries (LMICs) is mostly caused by <i>Salmonella typhi</i>. The development of antibacterial therapies is drawn to the enzyme 3-dehydroquinate dehydratase type 1 (DHQ1), which is an important enzyme of the shikimate pathway in bacteria. Its role in the synthesis of chorismite, a natural precursor of the route leading to the synthesis of aromatic amino acids, validates it as an important therapeutic target. Therefore, the current work is based on the investigation of bioactive compounds that inhibit the crucial DHQ1 enzyme. Structure-based virtual screening (SBVS) and docking studies were done using AutoDock Vina tools to retrieve potential hit molecules with high binding affinity against DHQ1 from several ligand repositories. From a total of 551 compounds, 206 were filtered out for physicochemical properties, the Lipinski rule and subjected to molecular docking. A total of 10 ligands were selected with the highest binding affinities ranging from -9.8 to -9.1 (kcal/mol), ie, Cabotegravir, Imatinib, Prulifloxacin, Limonin, Silibinin, Atovaquone, Betamethasone Valerate, GSK1324726A, Isavuconazole, and Raltegravir. In addition, ADMET analysis, bioactivity prediction, and pharmacokinetic property prediction were carried out as prediction of activity spectra for substance (PASS) analysis to further prioritize the best inhibitory compound. Molecular dynamic simulations were used to verify the stability of the protein \"DHQ1\" docked with 3 hit ligands Cabotegravir, Limonin, and Silibinin by calculating root mean square fluctuation (RMSF), root mean square deviation (RMSD), radius of gyration (Rg), H-bond interaction, and molecular mechanics/generalized Born surface area (MM/GBSA) scores. Based on molecular dynamic (MD) simulations, it is reported that Cabotegravir and Limonin showed the optimal binding features with bacterial DHQ1 enzyme and can be deemed as a repurposed drug candidate as compared to Silibinin, which did not show favorable interactions with DHQ1 and needs to be explored further against typhoidal <i>Salmonella</i>.</p>","PeriodicalId":9065,"journal":{"name":"Bioinformatics and Biology Insights","volume":"20 ","pages":"11779322261418889"},"PeriodicalIF":2.0,"publicationDate":"2026-06-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13247285/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148216353","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Web-Based Graphical User Interface Design Integrating MATLAB Server for the Mathematical Model of Human Cardiovascular-Respiratory System.","authors":"Jean Marie Ntaganda, Solange Mukeshimana, Marcel Ndengo, Theogene Bizimungu, Dominique Harelimana, Celestin Kurujyibwami, Marie Aimee Uwineza, Jean de Dieu Maniraguha, Lydie Mpinganzima","doi":"10.1177/11779322261438259","DOIUrl":"10.1177/11779322261438259","url":null,"abstract":"<p><p>A mathematical model of the human cardiovascular-respiratory system is proposed in several MATLAB graphical user interface (GUI) designs. However, it is observed that the latter are not web-based designed, making it hard for users to use them anywhere. This article proposes a web-based GUI design that integrates MATLAB server as a computational tool, worldwide accessible by users, upon subscription. The proposed web-based GUI design has been tested using a small sample of customers, males and females having different attributes relative to the model. After testing across normal and abnormal medical conditions in Rwanda, our findings demonstrate that the proposed web-based GUI design is highly efficient. This suggests robust usability, indicating that users worldwide can securely utilize it to assess pressures on the human cardiovascular-respiratory system.</p>","PeriodicalId":9065,"journal":{"name":"Bioinformatics and Biology Insights","volume":"20 ","pages":"11779322261438259"},"PeriodicalIF":2.0,"publicationDate":"2026-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13243759/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148209951","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Listeria Genome Identification Using DNABERT Embedding With LightGBM and SHAP-Based Explainable Classification.","authors":"Sajeev Ram Arumugam, Ananth J P, Sankar Ganesh Karuppasamy, V Senthil Murugan, Wulfran Fendzi Mbasso, Ambe Harrison","doi":"10.1177/11779322261457840","DOIUrl":"10.1177/11779322261457840","url":null,"abstract":"<p><p>Prompt and accurate identification of Listeria monocytogenes at the whole-genome level is essential for food safety surveillance and outbreak prevention, yet existing culture-based, PCR, and next-generation sequencing (NGS) workflows are either slow, labor-intensive, or rely on opaque machine learning models with limited interpretability. This study proposes an explainable genomic classification framework that couples transformer-based DNA embeddings with gradient boosting to distinguish Listeria from non-Listeria genomes. A curated dataset of 700 complete bacterial genomes (350 L. monocytogenes and 350 biologically related non-Listeria genomes) was assembled from the NCBI Assembly database and rigorously filtered to retain high-quality assemblies. Each genome was tokenized into overlapping 6-mers and encoded using the pretrained DNABERT model to obtain contextual genome-level embeddings, which were then classified with a LightGBM classifier. Model decisions were interpreted using SHapley Additive exPlanations (SHAP) to quantify the contribution of individual embedding dimensions and associated k-mer patterns to each prediction. After final verification of the confusion matrix, the proposed DNABERT + LightGBM + SHAP pipeline correctly classified 335/350 Listeria genomes and 330/350 non-Listeria genomes, corresponding to 665/700 correct classifications and a corrected accuracy of 95.00%. The same evaluation yielded precision of 94.37%, recall of 95.71%, F1-score of 95.03%, and an AUC of 0.9976. Comparative experiments show that the framework consistently outperforms conventional k-mer-based Random Forest, TF-IDF + SVM, CNN, XGBoost, and DNABERT + Logistic Regression baselines across all metrics. Beyond its strong predictive performance, SHAP-based analysis reveals discriminative sequence patterns that may correspond to putative genomic signatures of Listeria. Overall, the proposed approach provides a high-performance, interpretable tool for genome-scale Listeria identification and offers a transferable template for explainable pathogen genomics in food safety and public health applications.</p>","PeriodicalId":9065,"journal":{"name":"Bioinformatics and Biology Insights","volume":"20 ","pages":"11779322261457840"},"PeriodicalIF":2.0,"publicationDate":"2026-05-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13223348/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148147866","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}