Gene expression patterns combined with network analysis identify hub genes associated with bladder cancer

• Published in: Computational biology and chemistry Vol. 56; pp. 71 - 83
• Main Authors: Bi, Dongbin, Ning, Hao, Liu, Shuai, Que, Xinxiang, Ding, Kejia
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 01.06.2015
• Subjects: Bayes Theorem; Bioprocess; Bladder cancer; Co-expression network; Gene Expression Profiling; Gene Expression Regulation, Neoplastic; Gene Ontology; Gene Regulatory Networks; Hub genes; Humans; Protein Interaction Maps; Transcription Factors - genetics; Transcription Factors - metabolism; Urinary Bladder - metabolism; Urinary Bladder - pathology; Urinary Bladder Neoplasms - diagnosis; Urinary Bladder Neoplasms - genetics; Urinary Bladder Neoplasms - metabolism; Hub genes; Bladder cancer; Co-expression network; Bioprocess
• ISSN: 1476-9271; 1476-928X
• DOI: 10.1016/j.compbiolchem.2015.04.001

[Display omitted] •We identified 332 global DEGs and DEGs in different stages and grades of BC.•Hub genes of complex networks were explored based on these DEGs.•UBE2C, ACTA2, FABP4, CKS2, FN1 and TOP2A were hub genes of networks.•ECM-receptor interaction and focal adhesion were the significant pathways. To explore molecular mechanisms of bladder cancer (BC), network strategy was used to find biomarkers for early detection and diagnosis. The differentially expressed genes (DEGs) between bladder carcinoma patients and normal subjects were screened using empirical Bayes method of the linear models for microarray data package. Co-expression networks were constructed by differentially co-expressed genes and links. Regulatory impact factors (RIF) metric was used to identify critical transcription factors (TFs). The protein–protein interaction (PPI) networks were constructed by the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) and clusters were obtained through molecular complex detection (MCODE) algorithm. Centralities analyses for complex networks were performed based on degree, stress and betweenness. Enrichment analyses were performed based on Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases. Co-expression networks and TFs (based on expression data of global DEGs and DEGs in different stages and grades) were identified. Hub genes of complex networks, such as UBE2C, ACTA2, FABP4, CKS2, FN1 and TOP2A, were also obtained according to analysis of degree. In gene enrichment analyses of global DEGs, cell adhesion, proteinaceous extracellular matrix and extracellular matrix structural constituent were top three GO terms. ECM-receptor interaction, focal adhesion, and cell cycle were significant pathways. Our results provide some potential underlying biomarkers of BC. However, further validation is required and deep studies are needed to elucidate the pathogenesis of BC.