br Microarray gene expression profiling and data processing br RNA
2.5. Microarray gene expression profiling and data processing
2.5.1. RNA preparation and quality control
Total RNA from cultured RKO Bortezomib was extracted by Trizol and purified using RNeasy RNA extraction kit. Quality control of extracted RNA was subsequently examined by both Thermo Nanodrop 200 and Agilent 2100 bioanalyzer with utilization of Agilent RNA 6000 Nano Kit. Quality RNA will be subject to microarray analysis until it meets the following standards: 1.7 < A260/A280 < 2.2 by Thermo Nanodrop 200 and RIN > = 7 and 28S/18S > 0.7 by Agilent 2100 bioanalyzer.
2.5.2. Microarray processing and data analysis
A total of 6 GeneChip microarrays (Aﬀymetrix 901838) were hy-bridized with 3 pairs of samples to determine gene expression profiles of the NC and KD RKO cell samples according to the manufacturer’s instructions. Briefly, qualified total RNA was firstly subject to poly(A) tailing (37 C for 15 min) and biotin ligation (25 C for 30 min) using FlashTag Biotin HSR labeling Kit before its hybridization with the mi-croarray gene chips (48 C, 60 rpm, 16–18 h) in GeneChip Hybridization Oven 645. After the hybridization, Genechip Hybridization Wash and Stain Kit was used to wash and stain the array chips on GeneChip Fluidics Station 450. Finally, microarrays were scanned by GeneChip scanner 3000 and dat and cel files were obtained using GCOS 1.1. Raw data expression was imported to R (www.r-project.org) and analyzed by the Bioconductor aﬀy package (www.bioconductor.org). Logarithmic (base 2) intensity measures were obtained by RMA. Intensity was converted to nonlogarithmic values and rescaled by ad-justing mean intensity on each array to 400. Cel files and RMA values were deposited on Gene Expression Omnibus (www.ncibi.nih.gov/geo/ ).
The gene expression profile was preprocessed using Limma package in Bioconductor and Aﬀymetrix annotation files. The Background cor-rection, quantile normalization and probe summarization of the mi-croarray data were performed using the Robust Multi-array average algorithm to obtain the gene expression matrix.
2.5.3. Identification of diﬀerential expressed genes (DEGs)
Limma package was used to normalize the microarray raw data, and genes with (log2fold change) > = 2 and P < 0.01 indicated that there is a statistically significant diﬀerence between the cancer tissues and normal controls. A total of 602 up-regulated and 765 down-regulated diﬀerentially expressed genes were identified when comparing shSNRPA1 silencing with control shRNA groups.
2.5.4. Enrichment analysis of DEGs
The on-line functional annotation tool, DAVID (www.abcc.ncifcrf. gov), was then used to perform the GO-BP functional enrichment ana-lysis for DEGs, with the threshold of P < 0.01. Pathway enrichment analysis was done using both KEGG (www.kegg.jp/Kegg/pathway. html) and Reactome (www.reactome.org) databases, P < 0.01 was selected as the threshold value.
2.5.5. Identification of genes associated with SNRPA1 from DEGs
The Tumor Suppressor Gene database (bioinfo.mc.vanderbilt.edu/ TSGene) and the Tumor-associated Gene database (blog.synopse.info/ tag/Database) were used to identify all known oncogenes or cancer suppressor genes from the DEGs, special focus was placed on those genes interacting with SNRPA1 and also involved in inhibiting the proliferation and promoting apoptosis of cancer cells.
2.5.6. Pathway and network analysis
STRING version 9.1 (www.string.embl.de) was used to search in-teraction associations of the proteins with the confidence score of > 0.9. Cytoscape software (www.cytoscape.org) was then used for performing visualization of the PPI network. The HUB nodes with the top 5 degrees in the PPI network were obtained. BioNet was used to
identify the PPI sub-network of DEGs with a false discovery rate of < 0.01. KEGG database was used to perform the pathway enrichment analysis of genes in the core PPI sub-network with a threshold value of P < 0.01.