Objective: To screen the differentially expressed genes in human colorectal cancer (CRC) tissue.
Methods: Affymetrix oligonucleotide microarrays HG-U133 representing 32,264 human genes including 19,308 known genes and 12,956 expressed sequence tags (ESTs) were used to detect the gene expressions of CRC tissue paired with normal mucosa tissue. The microarray findings were confirmed by real-time quantitative reverse transcriptase-polymerase chain reaction (FQ-PCR). The gene expression profiles were analyzed by intersection and complement, rank sum test and t test.
Results: Totally 3,125 genes and ESTs expressed differentially were detected in normal and cancer tissues, consisting of 974 up-regulated and 2,151 down-regulated genes with 247 ESTs present in CRC tissue and absent in normal mucosa and 162 ESTs absent in CRC tissue but present in normal mucosa. A percent of 80.1% of the differentially expressed genes were not reported in the literatures.
Conclusion: The strategy of data mining provides a foundation for filtering molecular markers and interpreting molecular carcinogenesis of CRC.