Introduction
DNA methylation at CpG nucleotide sites in eukaryotes is a key epigenetic mark that helps regulate gene expression. Specific changes in CpG methylation occur in many human cancers, making them a promising biomarker for early cancer detection. However, existing assays can be costly, lack specificity to regions of interest and often provide only semi-quantitative estimates of the methylation fraction. Here, we present new tools to address these challenges including the use of a targeted methylome panel to decrease costs associated with next generation sequencing (NGS), methylation controls to calibrate quantitative assays and UMIs for accurate deduplication in low-diversity samples.

Experimental Procedures
Genomic DNA (gDNA) was prepared for sequencing using the Twist Methylation Detection System consisting of enzymatic methyl conversion library generation and hybrid capture using the Twist Human Methylome Panel. Twist CpG methylation specific controls were spiked into gDNA with different methylation rates prior to library preparation and taken through the Twist Methylation Detection System using homologous panels to demonstrate how these controls can be used to calibrate detection assays. Additionally, libraries were generated using cell free DNA (cfDNA) and either conventional or UMI-containing adapters during the library preparation ligation step to investigate the impact on quantitative detection and total unique coverage.

Data
Using the Twist Human Methylome Panel at 150x raw coverage achieves uniform coverage with low off-bait Picard metrics. 6.59 million CpG sites were detected using a minimum depth of 10X.

Target capture with the Human Methylome Panel allows for informative CpG calling of up to 82 samples on a single Illumina Novaseq S4 flowcell. In contrast, performing traditional whole genome bisulfite sequencing (WGBS) without target enrichment, only 3 samples could be run on the same flowcell. The Twist CpG methylation specific controls are constructed of 48 unique contrived sequences that contain a total of 8 different levels of methylation, ranging from 100% to 0%. Including these controls allows for quantitation of methylation levels in the experimental samples qualification of the enzymatic conversion process.

Conclusions
Our study uses several new products found in the Twist methylation detection portfolio that interrogate genome-wide methylation patterns for various applications. Not only have we found ways to better control assays, but we have also determined methods to account for lower costs compared to WGBS.