Fig. 1

Overview of the proposed method. The cell line and cancer gene expression are introduced into our experiment. All of gene expression data are normalized us by Z-score. Differentially expressed(DE) genes are introduced as the genes to be optimized. Gene annotation data obtain the structure information using the dRW algorithm. Then, the normalized gene expression data and gene annotations contained the GO structure information will be introduced into the gene set annotation optimization algorithm. Gene set annotation optimization: gene sets divided by the clustering algorithm will own similar annotations using gene expression data. Finally, the proposed method and the state-of-the-art methods are evaluated using three metrics