The GSE53819 microarray dataset was downloaded from the Gene Expression Omnibus database (http://www.ncbi.nlm.nih.gov/geo/ ), which is the largest open database of gene expression data (15 (link)). The data set used in the present study consisted of 18 samples of NPC primary tumor tissue and 18 control samples of normal nasopharyngeal tissue, based on the GPL6480 Agilent-014850 Whole Human Genome Microarray 4×44 K G4112F platform (Agilent Technologies, Inc., Santa Clara, CA, USA).
According to the platform, all probe numbers in the microarray data were mapped to their corresponding gene names. Regarding the genes corresponding to several probes, the average expression values of these probes were calculated to determine the expression value of the gene. Subsequently, the skewed distribution of data was converted into a normal distribution using a log 2 transformation, followed by normalization using the Median method (16 (link)). The Linear Models for Microarray Analysis package (http://www.bioconductor.org/packages/release/bioc/html/limma.html ) (17 (link)) in R language was used to screen for the DEGs between the NPC and control tissue samples. Multiple testing correction (18 (link)) was also performed using the Benjamini-Hochberg method (19 (link)). |Log fold change|>1 and false discovery rate <0.05 were set as the strict cutoffs for DEG identification.
According to the platform, all probe numbers in the microarray data were mapped to their corresponding gene names. Regarding the genes corresponding to several probes, the average expression values of these probes were calculated to determine the expression value of the gene. Subsequently, the skewed distribution of data was converted into a normal distribution using a log 2 transformation, followed by normalization using the Median method (16 (link)). The Linear Models for Microarray Analysis package (