Total RNA was extracted from normal controls and cancer patients from cervix, endometrium and vulvar tissues and was hybridized on Affymetrix HG133 A 2.0 microarray chips corresponding to more than 14,500 uniquely represented genes (NetAffx 32). A total of 35 samples were used to identify potential biomarkers and signatures in each type of cancer. The data were analyzed with the R language (version 3.0.2) and bioconductor package (version 2.13) [55 (link)]. The RMA algorithm [56 (link)] and log2 transformation were used for background correction and normalization of the data. For those genes that were represented in more than one probe, only probes with the highest average value across all arrays were kept. In order to identify differentially expressed genes, we performed a Student's t-test in unlogged data between normal and cancer tissues, and those genes with a value of p < 0.05 and a fold change (±) greater than 1.5, were considered significant. The microarray data (GSE63678) were submitted to GEO [57 ].
Free full text: Click here
