https://orcid.org/0000-0003-2505-5422
Figures
Abstract
Mastitis in dairy cattle is a serious issue that affects not just the animals but also has broad social, cultural, economic, and human consequences. It does in a wide variety of ways and the most remarkable of which are reduced milk yield and produce poor milk quality. This study takes an approach of bioinformatics to track down new targets and biomarkers which can be used to diagnose the clinical and subclinical forms of mastitis and at the same time find the way to treat and manage the disease. Comparing genes that express at a different level and the protein network, we identified three key genes (CDKN1A, FKBP5 and SLC7A5) and pathways that mastitis includes both in clinical and subclinical form. In functional term, multicellular organismal process regulation, cell population proliferation, protein binding are identified as critical biological processes. Additionally, machine learning algorithms applied to validate the identified candidate biomarkers. Potential repurposing drug targets are identified based on the commonly selected differentially expressed genes. This integrative approach not only provides insights into the molecular mechanisms underlying mastitis but also offers a robust framework for developing targeted therapies and diagnostic tools, ultimately contributing to better herd health and productivity. The findings from this study pave the way for precision veterinary medicine, with the ability to decrease the impact of the economic burden of mastitis on the dairy industry.
Citation: Auwul MR (2026) Bioinformatics approach to identify potential biomarker and drug target for the clinical and subclinical mastitis disease in dairy cattle. PLoS One 21(5): e0349172. https://doi.org/10.1371/journal.pone.0349172
Editor: Manasa Varra, Sri Venkateswara Veterinary University, INDIA
Received: August 4, 2025; Accepted: April 25, 2026; Published: May 13, 2026
Copyright: © 2026 Md. Rabiul Auwul. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All data utilized in this manuscript are available from NCBI repositories with the accession numbers, GSE15025, GSE24217, GSE24560, GSE75379, GSE159286, and GSE149856 and GSE51856. These datasets and the accompanying demo R code are available online in the GitHub repository, https://github.com/MdRabiulAuwul/Mastitis-Data.
Funding: This study was supported by the Gazipur Agricultural University (GAU) Innovation Project (2023–2024 FY to M.R.A.).
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Mastitis is one of worldwide and economically important diseases facing the dairy cattle industry. Primarily because bacteria infect the mammary glands leading resulting decrease in milk production as well as changes in its composition, thereby resulting into substantial economic losses to farmers growing them. Globally more than 80% of total world’s cattle milk comes from various countries [1]. Mastitis has become an endemic issue within global dairy sector as well as some cases assaulting cows, goats and buffaloes among others with mastitis affecting these breeds too. The illness can typically be grouped into two main types namely, clinical mastitis which shows visible signs of inflammation like swollen red painful udders containing abnormal looking fluids; subclinical mastitis where there aren’t any notable clinical signs but affects both quality and quantity accordingly. Several reasons lead to mastitis prevalence namely: poor management practices, limited veterinary services, significant environmental contamination and incorrect milking method as well lack of awareness amongst farmers.
Early identification and prompt intervention for mastitis are indispensable in minimizing its effect on the volume as well as quality of milk. The present diagnostic systems depend on somatic cell count (SCC) as well California Mastitis Test (CMT), which serves to measure inflammation indirectly from milk specimens [2,3]. The traditional method for diagnosis and treatment of mastitis included standard bacteriological techniques, which take time and often do not recognize the aetiological agents in subclinical cases. Besides, the treatment of mastitis has become more difficult due to the emergence of antibiotic-resistant strains of bacteria. Therefore, new biomarkers for this disease urgently needed to identify that are effective in early diagnosis and possible drug targets to help design novel therapeutic interventions within veterinary and animal science areas. The use of bioinformatics and machine learning methods is significant in finding potential biomarkers for mastitis decease. Advanced technologies like microarray and RNA sequencing (RNA-seq) allow for rapid and comprehensive analysis of gene activity across the entire genome. Technology have made a lot of progress in recent times opening new avenues for critical care medicine at precision levels to more effectively understand how different biological processes function on the molecular scale [4,5]. There have been several research identified potential biomarkers for mastitis disease [6–8].
There have been several research is going on finding the biomarkers and drug targets [9–12]and so the study progressing to discover the genes linked to mastitis in dairy cattle have combined the use of microarray meta-analysis [13] and RNA-seq data [14] together with some machine-learning algorithms. Several studies in the literature have focused on discovering biomarker genes for either clinical nor subclinical mastitis [15,16], but no one shown the clinical and subclinical interaction with drug targets identification and machine learning validation. In this study, we identified the significant biomarkers and pathways between the clinical and subclinical mastitis disease as well as potential drug targets and repurpose existing drugs detected for the early treatment of subclinical mastitis together with clinical mastitis disease with machine-learning (ML) validation approaches. A schematic workflow for analyzing differentially expressed genes (DEGs) using clinical and subclinical mastitis data in this study presented in Fig 1. The identified biomarkers and drug targets could greatly improve mastitis disease management and deepen our understanding of the condition.
This study began with preprocessing microarray and RNA-seq count data for both clinical and subclinical mastitis and analyzed DEGs with machine learning validation and performed downstream analyses.
2. Methods and materials
2.1. Data processing and differential expression analysis using combinatorial statistical methods
The RNA-Sequencing and microarray-based gene expression transcriptomics datasets for mastitis disease for both clinical and subclinical were collected from the NCBI Gene Expression Omnibus (GEO). Needed to preprocess these datasets before doing necessary analysis, excluding lowly expressed genes, impute missing values and log2 transformation with quantile normalization. and to remove the technical variability across studies, batch effects reduction was applied to facilitate robust gene expression analysis [14]. We used removeBatchEffect function to adjust the batch effects of each dataset [14]. Datasets were analyzed separately for identification of differentially expressed genes (DEGs). The DESeq2 approach [17] for RNA count data and limma [18] for microarray datasets have been applied to identify the DEGs. We considered only those common DEGs in all datasets with satisfying the statistical threshold parameters, adjusted p-value <0.05 along with [19,20].
2.2. Validation of the selected genes with machine learning algorithms
The identified genes have been validated with four popular supervised machine leaning (ML) classification algorithms, support vector machine (SVM) [21], naïve bayes (NB) [22] linear discriminant analysis (LDA) [23] and random forest (RF) [24]. We validated the identified biomarkers in two ways, firstly used internal validation process and for that we randomly select one dataset for train set and another dataset for test set each of with only commonly identified genes, secondly an external dataset used as test data with commonly identified gene. The accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and detection rate (DR) will be analyzed to check the validity of the selected genes [19].
2.3. Protein-protein interaction network analysis
STRING- a widespread web-based tool utilized for analyzing Protein-protein interaction (PPI) [25] and the network have been visualized with Cytoscape [26]. A PPI network is provided by STRING which represents the functional and physical relationships among some of the differentially expressed genes (DEGs) when their protein counterparts are used as input. In this analysis, a PPI network have been constructed using information from STRING, and further visualized and analyzed through the Cytoscape open-source platform. The hub proteins of the PPI network determined with the help of maximum degree connectivity scores in this analysis via the CytoHubba plugin [27] in Cytoscape. The higher the values of degree connectivity for nodes, more edges are attached in those hub proteins [19].
2.4. Gene ontology and pathway enrichment analysis
We explored the roles and signaling pathways of commonly differentially expressed genes by conducting Gene Ontology (GO) as well as pathway enrichment studies with g:Profiler, a well-known gene set enrichment web tool [28]. The GO terms cover three main categories namely, biological processes (BP), molecular functions (MF) and cellular components (CC). For pathway analysis, this study relied on the KEGG (Kyoto Encyclopedia of Genes and Genomes) databases [29]. We considered GO terms and pathways as significantly enriched when their adjusted p-value was below 0.05 [19,30].
2.5. Drug-gene interaction analysis
The drug-gene interaction analysis comprised of a widespread DSigDB (version 1.0) database for candidate drug identification which is highly related to selected DEGs [31]. The mastitis biomarker genes and pathways we found in cows are well-characterized human orthologs, so we tapped DSigDB to pick out a handful of promising repurposed drugs. Even though DSigDB is built for human genes, we mapped cattle gene equivalents to find drugs that could be repurposed for immune pathways, bridging the gap until cattle-specific resources expand. The DSigDB is a freely accessible web-based resource with detailed drug and comprehensive drug target information. DSigDB gene sets provide seamless integration with GSEA software, enabling researchers to connect gene expressions with various drugs and compounds. The database includes 22,527 gene sets, covering 17,389 unique drugs linking to 19,531 genes, making it a valuable resource for drug regeneration and translational research. The network of DEGs and the candidate drugs identified by the DSigDB database is implemented via Enrichr [32].
3. Results
3.1. Differential expression analysis
The RNA-Sequencing transcriptomics and microarray gene expression datasets for both clinical and subclinical mastitis disease have been collected from the NCBI Gene Expression Omnibus (GEO) [33]. Table 1 described the data used in this study. After collecting datasets from GEO, we keep only those genes whose sum of all CPM value more than 2 for the RNA-seq count datasets. The differentially expressed genes have been identified with limma for microarray datasets and DESeq2 for RNA-seq count datasets. The volcano plots for each dataset clearly indicated the presence of upregulated (red color) and downregulated (blue color) genes (Figs 2C, S1).
A) 38 common genes identified among clinical (Microarray and RNA count) and Subclinical mastitis datasets. Three genes (CDKN1A, FKBP5 and SLC7A5) have been identified as common among all the datasets; B) heatmap plot for the common genes in GSE75379 dataset. The graph indicated the presence of differentially expression in diseases groups than control groups; C) volcano plots for the GSE75379 and D) GSE149856 datasets with indicating the presence of upregulated (red color) and downregulated (blue color) genes. The gray color indicated the equally expressed genes.
3.2. Detection of shared differentially expressed genes
We selected our target genes those are common between microarray clinical mastitis datasets and RNA-seq count subclinical mastitis data or RNA-seq count clinical mastitis datasets and RNA-seq count subclinical mastitis data (Fig 2A). We finally selected 38 genes between clinical and subclinical mastitis datasets. These genes are used for further analyses. Three genes (CDKN1A, FKBP5 and SLC7A5) are identified as common among all clinical and subclinical mastitis datasets. The heatmap plot using pheatmap r package in Fig 2B indicated the differentially expression patters of the 38 genes over different sample. These genes are highly expressed in treatment groups then control groups.
3.3. Validation of the selected genes with machine learning algorithms
The validation of the identified 38 genes, we applied four classification algorithms, support vector machine (SVM), linear discriminant analysis (LDA), naïve bayes (NB) and random forest (RF). For internal validation, we used one RNA-seq count dataset of clinical and one for subclinical mastitis for this analysis (randomly select GSE159286 as training sets and GSE149856 as test set). The identified available common genes have been used in both training and test data for machine learning classifiers. The mastitis or control groups have been used in both dataset as the classes for classification. The results of this ML analyses with different matrices shown in Table 2 and in Fig 3 concluded that, the identification of the common genes among different datasets are valid, since almost 90% accuracy carried for all classification algorithms. For external validation, GSE159286 was used as the training dataset and the external GSE51856 as the independent test dataset. GSE51856 is an RNA seq study of bovine milk and blood monocytes that examines the host response to Streptococcus uberis, a major cause of mastitis in dairy cows. In this work, only the milk samples from GSE51856 were analyzed, comprising 44 samples in total, with 25 healthy controls and 19 infected cases. The results of the machine learning analyses using different datasets, as presented in S1 Table and S2 Fig, confirm the validity of identifying common genes across datasets. This is evidenced by the consistently high accuracy of nearly 90% across all classification algorithms. The accuracy, sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV) and detection rate (DR)results validate the identified genes among the datasets.
We used randomly selected dataset, GSE159286 as training set and GSE149856 as test set. We used 10-fold cross validation across four classifiers (SVM, LDA, NB and RF) and find almost 90% accuracy for all classifiers that indicated the validity of our commonly identified genes among clinical and subclinical mastitis datasets.
3.4. PPI network analysis with hub genes detection
To identify the signaling biomolecules, we analyzed PPI network around the proteins encoded by common 38 DEGs (Fig 4). The topological analysis identified 10 hub proteins (i.e., ACTB, ALB, IL6, IL1B, STAT3, TLR4, PTGS2, APOE, CXCL8, LOC534578/ VCAM1, CDKN1A, FKBP5 and SLC7A5) based on the top degree connectivity. In Fig 4, the larger the size of the nodes indicates highly connective genes, and the octagons indicates those genes that’s are present both in clinical and subclinical mastitis. The overview of these 10 hub genes briefly described in Table 3.
The PPI network analysis indicated the interconnection among the identified common genes. The larger the sizes of the octagons indicated as highly connected genes. Ten hub proteins (ACTB, ALB, IL6, IL1B, STAT3, TLR4, PTGS2, APOE, CXCL8, LOC534578, CDKN1A, FKBP5, and SLC7A5) were identified based on the highest degree connectivity.
3.5. Functional annotation analysis
We performed gene set enrichment analysis to get a clearer picture about the common differentially expressed genes (DEGs) in mastitis. Our findings revealed these genes play a significant role in biological processes, including the regulation of multicellular organismal processes, “cell population proliferation”, “response to wounding”, “intracellular signaling pathways”, and the “regulation of epithelial cell apoptosis”. In terms of molecular function, the DEGs were notably associated with the “protein binding” activity. Additionally, these genes were significantly enriched in the cellular component known as the “fibrinogen complex”. A summary of the top Gene Ontology terms is presented in Fig 5.
This figure shown the important enriched terms for each Biological Process (BP), Molecular Function (MF) and Cellular Component (CC) for the 38 commonly identified genes among clinical and subclinical mastitis datasets. The color intensity indicated the enrichment significance of the connected genes across the GO terms.
We performed a pathway analysis to pinpoint the pathways that were dysregulated and enriched with common DEGs. Our analysis identified several pathways including “Hepatitis B”, “Toxoplasmosis”, “FoxO signaling pathway”, “Inflammatory bowel disease”, “Measles”, “Malaria” and “HIF-1 signaling pathway”. The top terms KEGG pathways are summarized in Fig 6.
This figure shown the significant pathways for the commonly identified genes among clinical and subclinical mastitis datasets. The color intensity indicated the significance of the connected genes across the pathways.
3.6. Drug prediction analysis
The statistically significant repurpose drugs have been identified via DSigDB database for the common DEGs between clinical and subclinical mastitis. In Table 4, we have summarized the top 10 significant drugs based on adjusted p-values, odd ratio and combined score that had interaction with common DEGs. The dexamethasone, cycloheximide, ACMC-20mvek, 5-Fluorouracil, simvastatin, aspirin, N-Acetyl-L-cysteine, tetradioxin are the significantly identified drugs in this study to treat subclinical and clinical mastitis disease.
4. Discussion
Clinical as well as subclinical mastitis in dairy cattle is a severe problem with wide-ranging social, cultural, economic, and human repercussions in addition to the animals. By successfully identifying potential new biomarkers for clinical and subclinical mastitis, it is possible to revolutionize mastitis diagnostics and management in dairy farming. Early detection of mastitis and the ability to differentiate between clinical and subclinical forms can help farmers implement targeted intervention strategies, reducing the economic losses associated with this disease. In this study, ML and system biology approaches implemented to identify common genes, their associate GO, pathways and therapeutic drug target for leveraging both clinical and subclinical mastitis disease.
Compared to earlier bioinformatics reports in mastitis research, which relied on the identification of genes by differential analysis, our approach is very different. Nevertheless, we have implemented methodologies, particularly gene expression analysis, which provides key gene [19]. Most of the previous reports detected the biomarker genes either from clinical mastitis or subclinical mastitis [8,14,41–43]. We used our proposed pipelines in both clinical and subclinical forms of mastitis datasets and tried to find out the biomarkers associated with both of the types. Some of the previous reports use only p-value to detect biomarkers. If we only utilize one method, such as the p-value, there is a chance that we will find some unsuitable genes as the biomarker [19]. And the p-value needs to adjust with some adjustment methods to control the increased risk of false positives (Type-I errors). We applied Benjamini-Hochberg adjusted p-value along with fold change values to detect biomarkers from both clinical and subclinical mastitis in dairy cattle. Unlike prior studies, we validated via machine learning classifiers. The performance of state-of-the-art methods was evaluated the classification performances in which obtained higher score in accuracy, specificity and sensitivity.
Among the identified 13 potential biomarkers for both clinical and subclinical mastitis diseases, the Cyclin Dependent Kinase Inhibitor 1A or p21 (CDKN1A) gene plays a significant role in the cell cycle and responding to DNA damage. Its expression varies in the mammary epithelium of lactating Holstein cows that produce exceptionally high or low milk protein and fat percentages [44]. FKBP5 (FK506 Binding Protein 5) involved in the regulation of the glucocorticoid receptor and stress response. Recent gene expression comparisons between uninfected quarters of healthy cows and those with mastitis also identified FKBP5 as a differentially expressed gene [45,34]. It may influence the inflammatory response and immune function during mastitis. SLC7A5 (Solute Carrier Family 7 Member 5) is an amino acid transporter involved in cellular uptake of essential amino acids. It may be a factor in cell proliferation and immune responses during mastitis [46]. The SLC7A5 transporter gene use for granule cell development [47] and that may play a significant role in mastitis tissue development.
Interleukin 6 (IL-6) is a cytokine that contributes to both inflammation and B cells maturation. Its expression is elevated in mastitis, playing a role in the inflammatory response and maintaining immune homeostasis. Monitoring IL-6 levels in milk can aid in the early detection of mastitis, which is particularly crucial in cases of subclinical inflammation [48]. ACTB, also known as beta-actin, is a gene common to human organism that is mostly expressed and thus can serve as a reference in some gene expression studies. The ALB gene encodes serum albumin which is in high concentration in human blood [49]. Serum albumin regulates osmotic pressure and transports small molecules such as fatty acids at the same time. It also has an involvement in mastitis inflammatory process.
For immune response towards infections IL1B is a pro-inflammatory cytokine which plays an important role. The inflammatory development of mastitis relies on this [42]. IL6 can activate transcription factor called STAT3 (Signal Transducer and Activator of Transcription 3). In mastitis it contributes to its development and reduces milk production by dairy cows [50]. Bacterial lipopolysaccharides are known by TLR4 (Toll-Like Receptor 4), which activates innate immune responses. This gene is a great candidate marker for mastitis resistance because it plays a role in neutrophil migration into and out of the mammary gland throughout the infection [51].
PTGS2 (Prostaglandin-Endoperoxide Synthase 2; COX-2) is an enzyme that transforms arachidonic acid into prostaglandins which are accountable in inflammatory reaction. This gene is downregulated during inflammation of mastitis [52]. Apolipoprotein E (APOE) is a protein engaged in lipids transport as well as immune reaction. This also hinted a possible role during immune response to mastitis and alteration of inflammation. CXCL8 (C-X-C motif chemokine ligand 8; IL-8) serves as the chemokine responsible for recruiting neutrophils. It has been shown previously that this gene stood out among the three highest ranked genes supposed to be associated with E. coli mastitis [13]. VCAM1 takes part in the gathering of leukocytes around inflammation sites. In addition, it has been detected in lymph nodes from subclinical mastitis dairy cows [53]. In this manner, these genes intervene in different dimensions of resistance against disease, inflammatory responses and cellular activities leading to occurrence as well as development of bovine mastitis.
The identified GO and KEGG pathway enrichments terms could lead to the emergence of novel therapeutic approaches and vaccines to combat mastitis effectively. The identified “Response to wounding” Go term emerges as the most critical factor in mastitis progression in cows, directly driving the acute inflammatory reaction to bacterial invasion [54]. By promoting the quick growth of immune cells like neutrophils and macrophages, which support the inflammatory response and tissue healing after infection, “cell population proliferation” plays a crucial part in the development of mastitis in cows [7]. The “Intracellular signaling pathways” in mastitis progression propagate critical signals that coordinate immune cell activation, cytokine production, and pathogen defense within mammary cells [55,14]. The FoxO and HIF-1 signaling pathways directly influence in mastitis through infected mammary tissues by regulating inflammation, oxidative stress, immune responses and cell survival [56,57].
The identified mastitis biomarker genes and pathways we found in cows are well-characterized human orthologs, so we used DSigDB to identify promising repurposed drugs. DSigDB lacks native bovine gene mappings, requiring manual ortholog conversion that may introduce inaccuracies in veterinary drug repurposing predictions. This study lacks experimental validation of the biomarkers or therapeutic claims including the data source variation with small size sample collected different time. We suggest carrying out more biological and clinical research to explore whether the candidate drugs could be repurposed to treat mastitis. While bioinformatics and systems biology have been invaluable in this research, it's important to validate these findings clinically before they can be used as reliable biomarkers. Large-scale transcriptomic studies covering clinical and subclinical mastitis cases in dairy cows will be carried out in the future through clinical partnerships with veterinary facilities.
5. Conclusions
This study demonstrated the power of bioinformatics in advancing our understanding and management of mastitis. Through comprehensive analysis of transcriptomic data, this study identified several candidate new biomarkers and potential drug targets that could revolutionize the diagnosis and treatment of both clinical and subclinical forms of this disease. Furthermore, this study highlighted how combining machine learning and bioinformatics helps pinpoint the most promising genes that might be key candidate biomarkers for both clinical and subclinical mastitis. The identified key biomarkers could be transformed into diagnostic methods, which would allow for early and accurate identification of mastitis in both clinical and subclinical form. Thus enhancing prognosis and lessening the financial load on dairy farming by early detection and controlling with treatments of mastitis in subclinical stage. In order to apply these discoveries in the real world further experimental validation and clinical trials remain important. Additionally, our study helps to improve health care for dairy cattle by supports developing strategies for early detection and effective treatment against udder infections thereby saving the farmers from heavy financial losses as well as boosting their milk production.
Supporting information
S1 Fig. Volcano plots for the A) GSE15025, B) GSE24217, C) GSE24560, D) GSE50685 and E) GSE159286 datasets.
This figure indicating the presence of upregulated (red color) and downregulated (blue color) genes in each dataset. The gray color indicated the equally expressed genes.
https://doi.org/10.1371/journal.pone.0349172.s001
(TIFF)
S2 Fig. Accuracy boxplot for four machine learning classifiers on external validation.
https://doi.org/10.1371/journal.pone.0349172.s002
(TIF)
S1 Table. Different matrices for ML classification algorithms based on External test data.
https://doi.org/10.1371/journal.pone.0349172.s003
(DOCX)
Acknowledgments
I acknowledge Dr. M. Nazmul Hoque (Dept. of GOP, GAU) for his valuable guidance in selecting this study's topic.
References
- 1.
Database, FAOSTAT Stat. In: FAOSTAT Food and Agriculture Organization of the United Nations. 2020.
- 2. Sun X, Zhao R, Wang N, Zhang J, Xiao B, Huang F, et al. Milk somatic cell count: From conventional microscope method to new biosensor-based method. Trends in Food Science & Technology. 2023;135:102–14.
- 3. Menzies P. Udder Health for Dairy Goats. Vet Clin North Am Food Anim Pract. 2021;37(1):149–74. pmid:33541696
- 4. Panahi B, Farhadian M, Hejazi MA. Systems biology approach identifies functional modules and regulatory hubs related to secondary metabolites accumulation after transition from autotrophic to heterotrophic growth condition in microalgae. PLoS One. 2020;15(2):e0225677. pmid:32084664
- 5. Bansal M, di Bernardo D. Inference of gene networks from temporal gene expression profiles. IET Syst Biol. 2007;1(5):306–12. pmid:17907680
- 6. Gorji AE, Roudbari Z, Sadeghi B, Javadmanesh A, Sadkowski T. Transcriptomic analysis on the promoter regions discover gene networks involving mastitis in cattle. Microb Pathog. 2019;137:103801. pmid:31618669
- 7. Bakhtiarizadeh MR, Mirzaei S, Norouzi M, Sheybani N, Vafaei Sadi MS. Identification of Gene Modules and Hub Genes Involved in Mastitis Development Using a Systems Biology Approach. Front Genet. 2020;11:722. pmid:32754201
- 8. Darang E, Pezeshkian Z, Mirhoseini SZ, Ghovvati S. Identification of Key Genes and Potential Pathways Associated with Mastitis Induced by E. coli. Biochem Genet. 2023;61(1):202–20. pmid:35834114
- 9. Zhang Y, Zhou Y, Zhou Y, Yu X, Shen X, Hong Y, et al. TheMarker: a comprehensive database of therapeutic biomarkers. Nucleic Acids Res. 2024;52(D1):D1450–64. pmid:37850638
- 10. Ochoa D, Hercules A, Carmona M, Suveges D, Gonzalez-Uriarte A, Malangone C, et al. Open Targets Platform: supporting systematic drug-target identification and prioritisation. Nucleic Acids Res. 2021;49(D1):D1302–10. pmid:33196847
- 11. Kumar RR, Kumar A, Chuang CH, Shaikh MO. Recent Advances and Emerging Trends in Cancer Biomarker Detection Technologies. Ind Eng Chem Res. 2023;62:5691–713.
- 12. Frisoni GB, Festari C, Massa F, Cotta Ramusino M, Orini S, Aarsland D, et al. European intersocietal recommendations for the biomarker-based diagnosis of neurocognitive disorders. Lancet Neurol. 2024;23(3):302–12. pmid:38365381
- 13. Sharifi S, Pakdel A, Ebrahimi M, Reecy JM, Fazeli Farsani S, Ebrahimie E. Integration of machine learning and meta-analysis identifies the transcriptomic bio-signature of mastitis disease in cattle. PLoS One. 2018;13(2):e0191227. pmid:29470489
- 14. Ghahramani N, Shodja J, Rafat S, Panahi B, Hasanpur K. Integrative Systems Biology Analysis Elucidates Mastitis Disease Underlying Functional Modules in Dairy Cattle. Front Genet. 2021;12.
- 15. Zwierzchowski G, Haxhiaj K, Wójcik R, Wishart DS, Ametaj BN. Identifying Predictive Biomarkers of Subclinical Mastitis in Dairy Cows through Urinary Metabotyping. Metabolites. 2024;14(4):205. pmid:38668333
- 16. O’Reilly EL, Viora L, Malcata F, Pepler PT, Zadoks R, Brady N, et al. Biomarker and proteome analysis of milk from dairy cows with clinical mastitis: Determining the effect of different bacterial pathogens on the response to infection. Res Vet Sci. 2024;172:105240. pmid:38608347
- 17. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550. pmid:25516281
- 18. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43(7):e47. pmid:25605792
- 19. Auwul MR, Rahman MR, Gov E, Shahjaman M, Moni MA. Bioinformatics and machine learning approach identifies potential drug targets and pathways in COVID-19. Brief Bioinform. 2021;22(5):bbab120. pmid:33839760
- 20. Ali Hossain M, Asa TA, Rabiul Auwul M, Aktaruzzaman M, Mahfizur Rahman M, Rahman MZ, et al. The pathogenetic influence of smoking on SARS-CoV-2 infection: Integrative transcriptome and regulomics analysis of lung epithelial cells. Comput Biol Med. 2023;159:106885. pmid:37084641
- 21.
Boser B, Guyon I, Vapnik V. A training algorithm for optimal margin classes. In: Proceedings of the 5th annual workshop on computational learning theory. Pittsburg, USA; 1992. p. 144–52.
- 22.
John G, Langley P. Estimating continuous distributions in Bayesian classifiers. In: Morgan K, editor. In: Besnard P, Hanks S (eds) Proceedings of the 17th conference on uncertainty in artificial intelligence. USA; 1995. p. 338–45.
- 23. Fisher RA. The use of multiple measurements in taxonomic problems. Ann Eugenics. 1936;7:179–88.
- 24.
Ho TK. Random decision forests. Proceedings of the International Conference on Document Analysis and Recognition, ICDAR. 1995;1: 278–82. https://doi.org/10.1109/ICDAR.1995.598994
- 25. Szklarczyk D, Morris JH, Cook H, Kuhn M, Wyder S, Simonovic M, et al. The STRING database in 2017: quality-controlled protein-protein association networks, made broadly accessible. Nucleic Acids Res. 2017;45(D1):D362–8. pmid:27924014
- 26. Smoot ME, Ono K, Ruscheinski J, Wang P-L, Ideker T. Cytoscape 2.8: new features for data integration and network visualization. Bioinformatics. 2011;27(3):431–2. pmid:21149340
- 27. Chin C-H, Chen S-H, Wu H-H, Ho C-W, Ko M-T, Lin C-Y. cytoHubba: identifying hub objects and sub-networks from complex interactome. BMC Syst Biol. 2014;8 Suppl 4(Suppl 4):S11. pmid:25521941
- 28. Kolberg L, Raudvere U, Kuzmin I, Adler P, Vilo J, Peterson H. g:Profiler-interoperable web service for functional enrichment analysis and gene identifier mapping (2023 update). Nucleic Acids Res. 2023;51(W1):W207–12. pmid:37144459
- 29. Kanehisa M, Goto S, Furumichi M, Tanabe M, Hirakawa M. KEGG for representation and analysis of molecular networks involving diseases and drugs. Nucleic Acids Res. 2010;38(Database issue):D355-60. pmid:19880382
- 30. Auwul MR, Zhang C, Rahman MR, Shahjaman M, Alyami SA, Moni MA. Network-based transcriptomic analysis identifies the genetic effect of COVID-19 to chronic kidney disease patients: A bioinformatics approach. Saudi J Biol Sci. 2021;28(10):5647–56. pmid:34127904
- 31. Yoo M, Shin J, Kim J, Ryall KA, Lee K, Lee S, et al. DSigDB: drug signatures database for gene set analysis. Bioinformatics. 2015;31(18):3069–71. pmid:25990557
- 32. Kuleshov MV, Jones MR, Rouillard AD, Fernandez NF, Duan Q, Wang Z, et al. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acids Res. 2016;44(W1):W90-7. pmid:27141961
- 33. Barrett T, Wilhite SE, Ledoux P, Evangelista C, Kim IF, Tomashevsky M, et al. NCBI GEO: archive for functional genomics data sets--update. Nucleic Acids Res. 2013;41(Database issue):D991-5. pmid:23193258
- 34. Mitterhuemer S, Petzl W, Krebs S, Mehne D, Klanner A, Wolf E, et al. Escherichia coli infection induces distinct local and systemic transcriptome responses in the mammary gland. BMC Genomics. 2010;11:138. pmid:20184744
- 35. Buitenhuis B, Røntved CM, Edwards SM, Ingvartsen KL, Sørensen P. In depth analysis of genes and pathways of the mammary gland involved in the pathogenesis of bovine Escherichia coli-mastitis. BMC Genomics. 2011;12:130. pmid:21352611
- 36. Brand B, Hartmann A, Repsilber D, Griesbeck-Zilch B, Wellnitz O, Kühn C, et al. Comparative expression profiling of E. coli and S. aureus inoculated primary mammary gland cells sampled from cows with different genetic predispositions for somatic cell score. Genet Sel Evol. 2011;43(1):24. pmid:21702919
- 37. Sipka A, Klaessig S, Duhamel GE, Swinkels J, Rainard P, Schukken Y. Impact of intramammary treatment on gene expression profiles in bovine Escherichia coli mastitis. PLoS One. 2014;9(1):e85579. pmid:24454893
- 38. Moyes KM, Sørensen P, Bionaz M. The Impact of Intramammary Escherichia coli Challenge on Liver and Mammary Transcriptome and Cross-Talk in Dairy Cows during Early Lactation Using RNAseq. PLoS One. 2016;11(6):e0157480. pmid:27336699
- 39. Cebron N, Maman S, Walachowski S, Gausserès B, Cunha P, Rainard P, et al. Th17-related mammary immunity, but not a high systemic Th1 immune response is associated with protection against E. coli mastitis. NPJ Vaccines. 2020;5(1):108. pmid:33298970
- 40. Saenz-de-Juano MD, Silvestrelli G, Bauersachs S, Ulbrich SE. Determining extracellular vesicles properties and miRNA cargo variability in bovine milk from healthy cows and cows undergoing subclinical mastitis. BMC Genomics. 2022;23(1):189. pmid:35255807
- 41. Li L, Chen X, Chen Z. Identification of Key Candidate Genes in Dairy Cow in Response to Escherichia coli Mastitis by Bioinformatical Analysis. Front Genet. 2019;10:1251. pmid:31921295
- 42. Brajnik Z, Ogorevc J. Candidate genes for mastitis resistance in dairy cattle: a data integration approach. J Anim Sci Biotechnol. 2023;14(1):10. pmid:36759924
- 43. Kirsanova E, Heringstad B, Lewandowska-Sabat A, Olsaker I. Identification of candidate genes affecting chronic subclinical mastitis in Norwegian Red cattle: combining genome-wide association study, topologically associated domains and pathway enrichment analysis. Anim Genet. 2020;51(1):22–31. pmid:31808564
- 44. Wu X, Huang J, Liu Y, Li H, Han B, Sun D. Bta-miR-106b regulates bovine mammary epithelial cell proliferation, cell cycle, and milk protein synthesis by targeting the CDKN1A gene. Genes (Basel). 2022;13:2308.
- 45. Jensen K, Günther J, Talbot R, Petzl W, Zerbe H, Schuberth H-J, et al. Escherichia coli- and Staphylococcus aureus-induced mastitis differentially modulate transcriptional responses in neighbouring uninfected bovine mammary gland quarters. BMC Genomics. 2013;14:36.
- 46. Chen Y, Ma Y, Ji Q, Yang X, Feng X, Yao R, et al. Intracellular Staphylococcus aureus Infection Decreases Milk Protein Synthesis by Preventing Amino Acid Uptake in Bovine Mammary Epithelial Cells. Front Vet Sci. 2021;8:756375. pmid:34869729
- 47. Sokolov AM, Holmberg JC, Feliciano DM. The amino acid transporter Slc7a5 regulates the mTOR pathway and is required for granule cell development. Hum Mol Genet. 2020;29(18):3003–13. pmid:32821949
- 48. Bochniarz M, Ziomek M, Szczubiał M, Dąbrowski R, Wochnik M, Kurek Ł, et al. Interleukin-6 as a Milk Marker of Clinical and Subclinical Intramammary Infections (IMI) in Cows Caused by Streptococcus spp. Animals (Basel). 2024;14(7):1100. pmid:38612339
- 49. Shi Y, Cai J, Shi C, Liu C, Zhou J, Li Z. Low serum albumin is associated with poor prognosis in patients receiving peritoneal dialysis treatment. J Healthc Eng. 2022;2022:1–6.
- 50. Khan MZ, Khan A, Xiao J, Ma Y, Ma J, Gao J, et al. Role of the JAK-STAT Pathway in Bovine Mastitis and Milk Production. Animals (Basel). 2020;10(11):2107. pmid:33202860
- 51. Panigrahi M, Kumar H, Nayak SS, Rajawat D, Parida S, Bhushan B, et al. Molecular characterization of CRBR2 fragment of TLR4 gene in association with mastitis in Vrindavani cattle. Microb Pathog. 2022;165:105483. pmid:35318070
- 52. Wang W, Li R, Ye T, Zhang X, Chen C, Liang A-X, et al. Preliminary Study on Gene Regulation and its Pathways in Chinese Holstein Cows with Clinical Mastitis Caused by Staphylococcus Aureus. J Vet Res. 2022;66(2):179–87. pmid:35892111
- 53. Chen Y, Yang W, Xu C. Expression of Vascular Cell Adhesion Molecule 1 (VCAM-1) in the Mammary Lymph Nodes of Cows with Subclinical Mastitis. J Vet Res. 2017;61(2):203–9. pmid:29978074
- 54. Oviedo-Boyso J, Valdez-Alarcón JJ, Cajero-Juárez M, Ochoa-Zarzosa A, López-Meza JE, Bravo-Patiño A, et al. Innate immune response of bovine mammary gland to pathogenic bacteria responsible for mastitis. J Infect. 2007;54(4):399–409. pmid:16882453
- 55. Islam MA, Takagi M, Fukuyama K, Komatsu R, Albarracin L, Nochi T, et al. Transcriptome Analysis of The Inflammatory Responses of Bovine Mammary Epithelial Cells: Exploring Immunomodulatory Target Genes for Bovine Mastitis. Pathogens. 2020;9(3):200. pmid:32182886
- 56. Lei Z, Ali I, Yang M, Yang C, Li Y, Li L. Non-esterified fatty acid-induced apoptosis in bovine granulosa cells via ROS-activated PI3K/AKT/FoxO1 pathway. Antioxidants. 2023;12:434.
- 57. Cai M, Fan W, Li X, Sun H, Dai L, Lei D, et al. The Regulation of Staphylococcus aureus-Induced Inflammatory Responses in Bovine Mammary Epithelial Cells. Front Vet Sci. 2021;8:683886. pmid:34136558