cell type: HeLa genotype/variation: Transfected with G392E Neuroserpin
Treatment protocol
Neuroserpin expression was induced in 'on' samples with 2 ug/ml doxycycline 48 hours before collection. 'off' samples remained untreated.
Growth protocol
250,000 Cells per well were grown in six well plates for 72 hours. The cells were cultured in DMEM supplemented with 10% v/v Tet-free approved FBS (Clontech), 200 µg/ml Geneticin and 100 µg/ml Hygromycin B (both selective antibiotics from Invitrogen, Paisley, UK), at 37°C and 5% v/v CO2 in a humidified incubator.
Extracted molecule
total RNA
Extraction protocol
Cells were collected by trypsinisation and pelleted at 700 g for 10 min at 4⁰C. RNA was isolated with Tri Reagent (Sigma). RNA amplification was performed using Ambion Total Prep kit (Ambion, UK)
Label
biotin
Label protocol
Sample was labelled with biotin as part of the Ambion Total Prep kit (Ambion, UK) protocol.
Hybridization protocol
Hybridization was carried out following the beadchip manufacturer's instructions (Illumina)
Scan protocol
Illumina iScan with iScan Control Software 1.3.10+
Description
G392E_on_22_11
Data processing
Raw data were pre-processed using R software (http://www.r-project.org) and the bioconductor (http://www.bioconductor.org) package lumi (Du et al., 2008). Probes were removed from the data set where the Illimuna detection p-value was greater than 0.01. The data were then transformed using the variance stabilisation transformation algorithm (VST) and normalised using the quantile normalisation method. Differentially expressed genes for each condition (wild-type neuroserpin ‘on’ versus wild-type neuroserpin ‘off’, G392E neuroserpin ‘on’ versus G392E neuroserpin ‘off and Δ neuroserpin ‘on’ versus Δ neuroserpin ‘off’’) were identified by fitting linear models with the bioconductor R package limma (Smyth, 2004). To account for multiple hypotheses testing, p-values were adjusted using the Benjamini & Hochberg False Discovery Rate (FDR) correction (Benjamini and Hochberg, 1995). In order to assess the gene expression profiles at a pathway level, Gene Set Enrichment analysis (Subramanian et al., 2005) was performed on the normalised gene expression values for each group. Gene Set Enrichment analysis uses a priori defined gene sets to determine pathway differences between two phenotypic states. Gene expression data is therefore evaluated at the pathway level rather than the single gene. Gene Set Enrichment analysis first ranks the genes according to their relative difference in gene expression. The ranked list is then compared to gene sets (pathways) and an enrichment score (ES) is calculated for each gene. When a gene is present in the gene set of interest, the running enrichment score is increased. If the gene is absent the running enrichment score is decreased. The enrichment statistic is the maximum deviation of the running enrichment score from zero. Statistical significance is assessed by performing a permutation test procedure. To account for the differences in gene set size, a normalised enrichment score (NES) is calculated. Estimated significance levels are also adjusted for multiple hypothesis testing. False Discovery Rate (FDR) is the estimated probability that a gene set with a given enrichment statistic represents a false-positive finding. A total of 639 pathways (gene sets) were analysed from Kyoto Encyclopedia of Genes and Genomes, Reactome, and BioCarta databases (http://www.genome.jp/kegg/pathway.html,http://www.reactome.org/ and http://www.biocarta.com/genes/index.asp respectively). Gene sets containing less than 15 and greater than 500 genes were filtered out, along with gene sets containing no genes from the input data set.
