Fig. 2

Flowchart for creating the combined matrices. Firstly, a parser reads the DNA methylation and RNA sequencing matrices in input (computed as described in Fig. 1), and sends the next elaborations to two distinct processes. A step is responsible of the creation of the full header of the combined matrix with all the genes, of both the DNA methylation and RNA sequencing. The other step takes the parsed sample IDs to modify the identification of the sample (TCGA barcode) deleting the details of the performed experiments. After the creation of the sample IDs, the join step follows: the initials matrices are joined on the modified IDs and the new rows of the matrix are created, including both gene expression and gene methylation quantity. The join defines the rows with the values of the two experiments (on which the join is made because the sample id is present in both input matrices), and also the rows with values of only one experiment (if the sample is not available in both input matrices of DNA methylation and RNA sequencing)