*2.6. Genome-Wide Analysis of Neighboring Intron and Exon*

The inclusion of a specific exon and the methylation status of its flanking intron were analyzed across the whole genome in order to investigate the potential local effects of intronic DNA methylation on the inclusion of alternative exons. We determined the bordering exon and intron pairs using the *GenomicRanges* package in R programming, and implemented a logistic regression analysis in quasi-Poisson distribution, with the independent variable as the intronic DNA methylation levels and the dependent variable as the exon usage. The logistic regression model was carried out in both DCM patients and controls, respectively, and both in intron-exon and exon-intron pairs, respectively. The regression model was adjusted by the distance between intron and exon in the pair. Additionally, we identified the regions with concomitant DEU and DMR (*p* value <0.05, respectively). The distinguished regions underwent gene ontology analysis using *GeneTrail2* (version 1.6) website [23,24].
