Whole human genome oligo microarrays 44 k

Gene expression analysis of NCI-H520 and Hep2 cells upon deadenylase silencing was performed at NIMGenetics (Madrid, Spain) by using SurePrint G3 Human GE 60 K Microarray and Whole Human Genome Oligo Microarrays 44 k, respectively (Agilent Technologies, Inc., Santa Clara, CA). Initial analysis of the microarray data was performed with cut-off fold-change value ± 2. To exclude any possible transcripts affected by the transfection procedure itself, we used non-Τarget shRNA (Sigma), pLKO.1 empty vector (Sigma) and wild type cells as controls of expression. The microarray data have been deposited in the Gene Expression Omnibus database (www.ncbi.nlm.nih.gov/geo, accession numbers GSE67536 and GSE67598). The upregulated genes arising from the deadenylases silencing were used to predict significantly enriched gene ontologies (GO). GO analysis was performed using GeneMANIA (www.genemania.org) on February 1st 2015 [28 (link)].

We analyzed publicly available gene expression data from Muggerud et al.[28 (link)] including tumor samples from 31 cases of pure DCIS, 36 pure invasive cancers, 42 cases of mixed diagnosis (invasive cancer with an in situ component) and 6 normal breast tissue samples. In this study, global gene expression was assayed using the Agilent Whole Human Genome Oligo Microarrays 44k. Available data on samples included clinical information, e.g. age at onset, and pathology data, e.g. tumor grade, and ER, PR, ki67 and HER2 amplification status. For each tumor sample, the relative expression of each gene (RAD51, TODRA, TPIP and E2F1), was normalized to the average expression of the same gene in the normal tissue samples. Correlation between expression levels of the different genes, and between expression levels and pathological and clinical information was performed using Pearson correlation for continuous variables and Spearman correlation and t-test for non-parametric comparisons (PASW Statistics 18).