Figure 2:

Quality control (QC) performed on each omic dataset. (A) Read QC using FastQC for WES data. (B) Read QC using FastQC for WGS data. (C) Read QC using FastQC for RNA-Seq data. For A, B, and C, the left panels correspond to the sequence quality plots, the x-axis representing the base position in the read and the y-axis the mean quality value; the right panels correspond to the per sequence quality score plots, the x-axis representing the mean quality score and the y-axis the number of reads. (D) QC of the RNA-Seq data after trimming. Left: Bar plot representing the percentage of reads uniquely mapped (“Uniquely mapped”), mapped to multiple loci (“Mapped to multiple loci” or “Mapped to too many loci” if the number of loci is >10), unmapped because the mapped reads’ proportion was too small (“Unmapped: too short”), unmapped because of too many mismatches (“Unmapped: mismatches”), or unmapped for other reasons (“Unmapped: other”). Middle: Cumulative bar plot representing the percentages of reads mapped, using RSeQC, at different locations in the genome (exons, introns, 5′ and 3′ untranslated transcribed region [UTR], intergenic regions, TSS, and TES). Right: Cumulative bar plot representing the cumulative percentages associated with the different read assignments using HTSeq (“Assigned”: reads assigned to 1 gene, “Ambiguous”: reads assigned to multiple overlapping genes, “Aligned not unique”: reads assigned to multiple non-overlapping genes, “No Feature”: reads assigned to none of the features). (E)  Left: Samples’ quality based on log median intensities. The x-axis and y-axis correspond to the median of log₂ methylated and unmethylated intensities, respectively. Right: Representation of the between-sample similarities based on the 2 first multidimensional scaling dimensions. (F) Histogram of the median detection P-value for each sample. (G) Distribution of the β-values for each sample before and after the filtering step (left and right panel, respectively).