A Method For Identifying Gene Functional Modules Based On Multidimensional Genome Data

The appearance of microarray makes us enable to measure the expression levels of tens of thousands of genes simultaneously, and it takes biology research a lot of help while also takes data analysis new puzzles. In those methods of selecting di?erentially expressed gene sets use gene expression pro?le, multivariate statistic test is a common and e?ective one. But, get the whole gene set expression level of every sample is helpful for individualized disease research, what promote the developing of method of data dimension reduction.In this paper, we conduct di?erentially expressed pathway(common prior gene set) analysis use gene expression pro?le after introduced some classic di?erentially expressed analysis. The key method of this paper is the method of data dimension reduction. After we introducing some classic method of data dimension reduction,we propose a new method that is more suitable for gene expression data based on an improvement of locality preserving project. Then, we take two real data set, and analysis them by our new method as well as those classic ones. After all of this,we take those results to compare with the result of one classic method of selecting di?erentially expressed pathway(gene set enrichment analysis), and then we get the feasible of the new method.