PURPOSE: The purpose in this study was to select key genes related to ovarian cancer. METHODS: The gene expression profiles of E-GEOD-6008, E-GEOD-26712, E-GEOD-27651, E-GEOD-14001 were obtained from ArrayExpress database (http://www.ebi.ac.uk/arrayexpress/). Following data recruitment and preprocessing, differentially expressed genes (DEGs) were characterized using Significance Analysis of Microarrays (SAM). Then, a differential expression network (DEN) was constructed using Cytoscape 2.1 software based on differential and non-differential interactions. Pathway analysis was performed based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database using Pathway Analysis with the nodes contained in the main DEN. Centrality analysis on the DEN was conducted to selected HUB genes. And last, western blot was performed on the selected genes in an independent sample set. RESULTS: A total of 370 samples (347 ovarian tumors and 23 controls) were selected. In all, 490 DEGs were obtained, which contained 59 upregulated and 431 downregulated genes. A DEN including 875 gene pairs (1028 nodes) was constructed. There were 7 pathways by analyzing the nodes contained in the main DEN. Five HUB genes were gained, and three (UBC, ELAVL1, SIRT1) were both HUB genes and disease genes. Meanwhile, SIRT1 and NEDD4 were downregulated genes. Verification experiments indicated that the expression of SIRT1 and ELAVL1 in the disease group and the normal group were significantly changed. CONCLUSIONS: This study showed that SIRT1 could be chosen as a potential biomarker for promoting detection of ovarian cancer, so as to further understand the molecular pathogenesis of this disease.
PURPOSE: The purpose in this study was to select key genes related to ovarian cancer. METHODS: The gene expression profiles of E-GEOD-6008, E-GEOD-26712, E-GEOD-27651, E-GEOD-14001 were obtained from ArrayExpress database (http://www.ebi.ac.uk/arrayexpress/). Following data recruitment and preprocessing, differentially expressed genes (DEGs) were characterized using Significance Analysis of Microarrays (SAM). Then, a differential expression network (DEN) was constructed using Cytoscape 2.1 software based on differential and non-differential interactions. Pathway analysis was performed based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database using Pathway Analysis with the nodes contained in the main DEN. Centrality analysis on the DEN was conducted to selected HUB genes. And last, western blot was performed on the selected genes in an independent sample set. RESULTS: A total of 370 samples (347 ovarian tumors and 23 controls) were selected. In all, 490 DEGs were obtained, which contained 59 upregulated and 431 downregulated genes. A DEN including 875 gene pairs (1028 nodes) was constructed. There were 7 pathways by analyzing the nodes contained in the main DEN. Five HUB genes were gained, and three (UBC, ELAVL1, SIRT1) were both HUB genes and disease genes. Meanwhile, SIRT1 and NEDD4 were downregulated genes. Verification experiments indicated that the expression of SIRT1 and ELAVL1 in the disease group and the normal group were significantly changed. CONCLUSIONS: This study showed that SIRT1 could be chosen as a potential biomarker for promoting detection of ovarian cancer, so as to further understand the molecular pathogenesis of this disease.