Single Nucleotide Analysis of Cytosine Methylation by Whole‐Genome Shotgun Bisulfite Sequencing

Abstract

Unlike other methods to assess methylation, whole-genome shotgun bisulfite sequencing is used to generate quantitative genome-wide methylation profiles at single-nucleotide resolution. As described in this unit, this method allows for the quantitative measurement of methylation at each cytosine base by treatment of genomic DNA with sodium bisulfite followed by sequencing and alignment of the reads to a reference genome.

Functional variation in allelic methylomes underscores a strong genetic contribution and reveals novel epigenetic alterations in the human epigenome

“…Both reads in a pair were trimmed of any low-quality sequence at their 3′ ends (with Phred scale score ≥30). Post-process read mappings were made as previously described [44], including clipping 3′ ends of overlapping read pairs in both forward and reverse strand mappings, filtering duplicate, low-mapping quality reads, read pairs not mapped at the expected distance based on the library insert size, as well as reads with more than 2% mismatches. Methylation calls of individual CpGs were extracted using Samtools in mpileup mode.…”

“…Reads were then mapped to these two reference genomes using the sample pipeline as described above [44], except that no mismatches were allowed during the alignment step in order to ensure that reads coming from a specific allele will map to the appropriate reference. An in-house software takes the genotypes and their positions and scans the alignment files to obtain the methylation states of the CpGs surrounding the alleles.…”

Global Analysis of DNA Methylation Variation in Adipose Tissue from Twins Reveals Links to Disease-Associated Variants in Distal Regulatory Elements

“…Each sample was sequenced on one lane of an Illumina HiSeq 2000 or 2500 system with the use of 100 bp paired-end sequencing, yielding on average 366 million total reads per sample. Reads were aligned to the bisulphite-converted reference genome with the Burrows-Wheeler Aligner; (1) clonal reads, (2) reads with a low-mapping quality score (<20), (3) reads with a more than 2% mismatch to the converted reference over the alignment length, (4) reads mapping to the forward and reverse strand of the bisulphite-converted genome, (5) read pairs not mapped to the expected distance according to the library insert size, and (6) read pairs mapping in the wrong direction were removed as described by Johnson et al 39 For avoiding potential biases in downstream analyses, the following WGBS-interrogated CpG sites were further filtered: sites not covered by at least three reads, sites overlapping a SNP (dbSNP 137), and sites overlapping DAC Blacklisted Regions or Duke Excluded Regions generated for the ENCODE project. The mean genome coverage was estimated to be~7-fold.…”

A Bayesian Approach for Analysis of Whole-Genome Bisulfite Sequencing Data Identifies Disease-Associated Changes in DNA Methylation

“…These fragments were then amplified by the Illumina cluster robot, and transferred to the HiSequation 2000 for sequencing. WGBS reads were aligned and filtered according to a previously published pipeline (see Johnson et al 2012, 2014). Briefly, reads were preprocessed by in silico conversion of C bases to T bases in read 1, and G bases to A bases in read 2, followed by clipping of the first base from each read.…”