Study On Identifying The Related Pathways And Key Genes With The Complex Diseases By Meta-Analysis Of Microarrays

The complex diseases are caused by the interaction of a number of genetic and environmental factors. With the improvement in microarray technology and its increasing use in the study of animal complex diseases, especially using human, mice, rats and other animals as a model of animal cancer studies, we have generated a lot of highly complex data sets. How we can eventually find out the true and reliable information useful for us from these massive data in order to discover the laws of molecular genetics inducing the complex diseases, which is becoming a major problem for us. Meta-analysis, a statistical approach that combines results from independent but related studies, is a relatively inexpensive option that has the potential to increase both the statistical power and generalizability of single-study analysis. Therefore, to get more comprehensive and reliable biological information which is related to one of the complex diseases (endometriosis), we have used meta-analysis method based on the existing and available DNA microarray data sets for this disease.To conduct a meta-analysis of different sources of data sets, we have used two chip data sets which are both related to endometriosis from the database of GEO. A novel meta-analysis we have used here with the package program of R is designed to perform rank product (RankProd) on lists of genes that are derived from different studies. We have also compared the result with the single analysis. The conclusion is that the result of RankProd is more sensitive and specific. Furthermore, we have discovered that there are two significantly related genes named THBS1 and THBS2 to the endometriosis with their pathway of extracellular matrix receptor interaction.In order to study how these two factors THBS1 and THBS2 participate in the regulation of reproductive diseases, which share a number of structural similarities, interact together and may be to some degree functionally redundant, we have carried out co-expression frequency meta-analysis through a web-based tool OncomineTM. Finally, there are many genes or proteins identified as being co-expressed with THBS1 and THBS2. Then, we use another web tool DAVID and other tools to conduct the gene ontology analysis and pathway analysis. We have discovered some significant pathways related to this disease and some novel partners of the related pathway of extracellular matrix receptor interaction. This method solves the limitations of meta-analysis based on only two datasets, and can conduct a meta-analysis of a wider range of array datasets with regard to specific gene sets for more comprehensive and accurate results.