DNA methylation is an epigenetic modification that plays an important role in regulating gene expression. There is evidence that the hypermethylation of promoter regions always cause gene silencing. However, how the methylation patterns of other regions in the genome, such as gene body and 3’UTR, affect gene expression is unknown. In this paper, we develop a differential-based method to analyze the relationship between the methylated change in different regions and that of corresponding gene expression. Through applying the method in five cancer datasets from the Synapse database, it is found that methylated regions with significant difference between cases and controls are almost uniformly distributed in the seven regions of the genome. Also, there are different effects on gene expression. For example, there are higher percentage positive relationships in 1stExon, gene body and 3’UTR than TSS1500 and TSS200. The functional analysis of genes with significant positive and negative correlation between DNA methylation and gene expression demonstrates the epigenetic mechanism of cancer-associated genes.