2.5 Genome wide methylome analysis
After visual inspection of the quality of DNA methylation profiles using
the Integrative Genomics Viewer (Thorvaldsdóttir et al., 2013), we
performed the analysis of the methylome at the genome’s scale. In order
to evaluate the global methylation level, metagene analysis was
performed using deepTools (version 2.0, Ramirez et al., 2016) command
“computeMatrix” to generate read abundance from all samples over
genomic regions: promoter, 5’UTR exons, coding exon, first intron,
internal introns (located in non-flanking regions of genes), last
intron, 3’UTR exon and Transcription End Site (TES). This matrix was
then used to create, using deepTools command “plotProfile”, a metagene
profile from 2kb upstream of the Transcription Start Site (TSS) to 2 kb
downstream of the Transcription End Site (TES). The same method was used
to generate a profile plot of the level of methylation across all
genomic regions.
The methylation profiles of the samples were studied using the R package
MethylKit (Akalin et al. 2012). The alignment BED files were first
converted into a tabular file suitable for the MethylKit package using
the methylextract2methylkit tool (version 0.1.0). In order to increase
the power of the statistical tests, the samples were filtered according
to read coverage. Bases that had less than 10X coverage and those that
had greater than 99.9th percentile coverage in each sample were filtered
out from the analysis to account for potential PCR bias.
Hierarchical clustering analysis and principal component analysis were
performed using the ”ClusterSamples” and ”PCASamples” functions,
respectively, of the Methylkit R package. These analyses were based on
similarities in the methylation patterns of the samples from each
salinity condition. A distance correlation matrix was generated with the
Pearson method and the clustering was performed using the Ward method.