Target-enriched enzymatic methyl sequencing

TEEM-Seq (target-enriched enzymatic methyl sequencing), a method that combines enzymatic methyl sequencing with a custom-designed hybridization capture bait set that can be scaled to reactions including large numbers of samples in any species for which a reference genome is available.DNA from 39 superb starlings (Lamprotornis superbus) and two house sparrows (Passer domesticus) was quantified using a Qubit 3.0 Fluorometer (Thermo Fisher Scientific), and 100-150 ng per sample was sheared to 250-275 bp on a Covaris ME220 and inspected on an Agilent 4200 TapeStation. The sheared DNA was then used as a template for libraries prepared with NEBNext Enzymatic Methyl-seq Kit (EM-seq) (New England BioLabs). Library yield was increased for myBaits in-solution biotinylated RNA capture of target gene regions (myBaits v4.01 Custom 1-20K 16 Reaction Kit (Daicel Arbor Biosciences) designed for 22 promoter and 12 complete exon ranges) through 12 total PCR cycles. We Qubit-quantified the libraries, inspected them on the TapeStation, and pooled 25 ng from each library into 1.2 microgram total for the pool of 48 libraries, including 5 starling and 2 house sparrow replicates. After hybridization reaction at 63C MyBaits hybridization beads bound to the pooled library-blocker mix were cleaned with three washes, and the resuspended bead-bound DNA was then amplified with 16 PCR cycles in a KAPA HiFi reaction using a 60C annealing temperature and a one minute extension step at 72C. The amplified capture pool was subsequently cleaned up with AMPure XP beads and sequenced at 2x150 bp with 5% PhiX spike-in in one lane (110 Gb) of an Illumina HiSeq 4000 at the Novogene Corporation, Sacramento, CA. Extra material from three of the EM-Seq libraries pre-capture were pooled and sequenced at 2x150 bp with 5% PhiX spike-in in 3/4th of a lane (82.5 Gb) on a HiSeq 4000 at Novogene with the same parameters to provide a whole-genome sequence and partial methylome analysis comparison to our targeted sequencing.NuGEN Ovation Methyl-Seq kits were used to prepare RRBS libraries (with added Promega unmethylated cl857 Sam7 lambda control spike-in of 1 ng per 80-120 ng of sample), which were sequenced in several pools of 16 samples per lane (20 Gb) at 1x100 bp reads with 10% PhiX spike-in on a HiSeq 2500 at the Cornell Institute of Biotechnology Genomics Core, Ithaca, NY. Eight of these RRBS samples were included in the present study, though one failed enzymatic conversion. Data from the remaining seven samples (from different sequencing pools) were used as a comparison to the TEEM-Seq and WGEM-Seq derived data generated from the same individuals.