The conflict of our findings and those of EMT are highlighted by the upregulation of genes associated with EMT in our microarray analysis, such as hepatoctye growth factor (HGF) (+9.6 fold).
Based on enriched GO terms and important superfamilies in 1,062 CIPs, we selected some candidate genes for real-time RT-PCR analysis to confirm our microarray analysis, such as glutathione S-transferase genes, cytochrome P450s, NBS-LRR, and chalcone biosynthetic related genes, etc.
Genes expected to be transcriptionally quiescent based on microarray analysis, such as PLL4 or PLL24, showed promoter activity that was relatively strong, and that was similar to genes that showed strong transcriptional signals such as PLL8 or PLL15, respectively.
Other genes that showed a similarly dramatic response upon microarray analysis, such as Sven_2720 encoding a putative transcriptional regulator fused to a uroporphyrinogen III synthase-like domain (homologous to SCO2958), were therefore also strong candidates for GlnR regulation.
Microarray analysis, such as the Affymetrix Human Genome Gene Chips, is an exciting development in breast cancer diagnostics that allows the expression of genes in tumors to be quantified using RNA retrieved from breast cancer biopsies.
Several statistical methods have been developed for different aspects of genomic microarray analysis, such as preprocessing (normalization), automatic detection of copy-number variations, and the analysis of Type I errors across genomic microarrays obtained from multiple experiments and samples.
Although an extensive microarray analysis of such model has not been published, we have been able to retrieve a dataset from the GEO database (GSE18033).
Gene Ontology has been widely used in genome research applications, and also for the validation of results obtained after a microarray analysis process, such as clustering or biclustering.
However, ROMA includes other sources of false positives, most importantly read-through transcripts into adjacent genes due to inefficient transcription termination in vitro, as well as ambiguities derived from impure protein preparation or the microarray analysis as such (65).
Because the abundance of each shRNA was measured by a single probe instead of using multiple ones as in many mRNA expression arrays, the noise level in the data was relatively high and classical microarray analysis methods such as t-statistics followed by FDR were not suitable to analyze this type of data.
However, the bottleneck is the construction of an expression matrix of TDFs (rows) per time points (columns) of HiCEP electrophoretic data due to the problem of imperfect alignment, though most microarray analysis uses such matrices as given data (e.g., [ 17, 18]).
