Identifikation von differentiell exprimierten Genen in soliden Tumoren verschiedener Herkunft

Studienziel: Identifikation von differentiell exprimierten Genen in soliden Tumoren verschiedener Herkunft
Software: Excel, SAM
Kriterien für die von Kanditatengenen: SAM q Wert < 25% und differentielle Expression in mindestens 10 verschiedenen Tumorgeweben
Ergebnisse: The majority of malignant tumors are solid tumors. Of these, cancer originating from epithelial cells are the most common. It is well known that most of the carcinomas harbour mutations in the same tumor suppressor genes like p53 and p16. Nevertheless, no common expression profile of solid tumors has been described so far. Therefore we were interested if genes differentially expressed in the majority of carcinomas could be identified using a meta-analysis approach. Complete data sets from the Stanford microarray database, the GEO database and the URLs described in the respective publications were downloaded for carcinomas of the prostate, breast, lung, ovary, colon, pancreas, stomach, bladder, liver, and kidney. All data available for download until April 2003 were included. Subsequently, all datasets were subjected to an expression analysis using SAM. In each experiment genes were scored as differentially expressed if the q-value was below 25%. Representative sequences of the probes of the different microarrays were obtained and compared to the Unigene build 155 using blastN. To obtain differentially expressed genes within the set of analysed carcinomas the number of observed differential expression in the experiments were counted. Differentially expression was assigned to genes if they were differentially expressed in at least eight experiments with tumors from different origin and have not been identified as differentially expressed within the counter direction, e.g. a gene was upregulated in eight but downregulated in none of the experiments. Using this approach we identified 100 genes as upregulated and 21 genes as downregulated in the carcinomas analysed. The genes represent various functional classes. Especially genes of the proteasome were found to be overexpressed. The most promising candidate genes are subjected to further validation. In summary, genes resembling a new source of potential diagnostic markers and targets for therapeutic intervention for the different types of carcinomas could be identified using a novel meta-analysis approach.