For detailed methods on sample collection and DNA isolation see Sepers et al., 2021. Several major improvements were made to the original epiGBS laboratory protocol (Boquete et al., 2020). In short, the genomic DNA was digested with the restriction enzymes MspI and NsII. A random three letter oligonucleotide (UMI) was placed in the adapter sequence (Moorsel, et al., 2019). Furthermore, a control nucleotide (CN, an un-methylated cytosine) was placed after the barcode followed by the sequence of the RE overhang (Moorsel, et al., 2019). The cytosines of the oligonucleotides adapter BA-I and adapter CO-I were 5-C methylated. The oligonucleotides of the opposite strands (adapter BA-II and adapter CO-II) contained un-methylated cytosines only and are 5-de-phosphorylated. After annealing the respective BA-I and BA-II and CO-I and CO-II adapter oligonucleotides and ligating them with the enzyme digested DNA fragment, only adapter 3 ends and fragment 5 ends ligate. The nick between adapter 5 ends and fragment 3 ends was repaired by using dNTPs that contained 5-meCs and that directly translated all 5-3 nucleotides starting from the nick. This resulted in fully methylated adapters that were ligated to the digested DNA fragment and a complementary 3-nucleotide short UMI sequence. As the four samples were pooled with 44 other samples, this resulted in a single, multiplexed sequencing library containing 48 barcoded samples. Library construction was conducted at the NIOO-KNAW. The final epiGBS2 library was sequenced on an Illumina HiSeq X Ten (150 bp from paired-end reads) by Novogene (Novogene (HK) Company Limited, Hong Kong, China) in 2019. The resulting epiGBS2 library was paired-end and directional, all four different bisulfite strands were present in the sequenced library: the original top strand (Watson), the complementary Watson strand, the original bottom strand (Crick), and the complementary Crick strand.
|Date made available||2021|