Integration of DNA methylation with gene expression

We integrated the 450K array and RNA-seq data by following analytical process: (1) identification of differentially-methylated and -expressed genes in 15 paired ESCC samples. The differentially-methylated genes were selected by a p-value < 0.05, and (2) testing of both differentially-methylated and differentially-expressed genes for a strong association. Pearson correlation analysis was used to identify their correlations. A negative correlation was considered significant for a Pearson correlation coefficient (PCC) < -0.5 and p-value < 0.05.

The DAVID Bioinformatics Tool (version 6.8) [63] was used to infer the potential biological processes of methylation-associated genes. Results with p-value<0.05 were considered as significant functional categories.