BACKGROUND AND OBJECTIVELung adenocarcinoma (AC) is the most common type of lung cancer, however, its mechanism of oncongenesis is still unknown. The aim of this study is to screen candidate genes of lung adenocarcinoma using bioinformatics strategy and elucidate the mechanism of lung adenocarcinoma.

METHODSTwo published microarray data (GSE7670 and GSE10072) was obtained from Gene Expression Omnibus (GEO). Significance analysis of microarrays was performed with the software dchip, and differential expression genes from dchip analysis were defined as "test gene set". Genes correlated with lung adenocarcinoma, obtained by data mining tools genecard and Fable were regarded as "train gene set". Finally, candidate genes of lung adenocarcinoma were screened by the tool "Toppgene".

RESULTSThree hundred and forty-four differential genes were defined as "test gene set", and 277 genes correlated with lung adenocarcinoma were regarded as "train gene set". Thirty-six candidate genes were screened out by Toppgene, among them, 21 genes had nearly no report in cancer. In the following QRT-PCR experiment, CD36, PMAIP1 and FABP4 were down-regulated expression in A549, which coincided with the gene chip.

CONCLUSIONIt is demonstrated that Toppgene is useful in identification of the candidate genes of lung adenocacinoma, which provides the proof for the discovery of the specific disease genes.