Pig and quail CpG methylation datasets from short and long read sequencing technologies
Reading CpG methylation across livestock and avian genomes is harder than it sounds. Different sequencing technologies disagree with each other — particularly across high-GC regions like CpG islands — and most labs lack a same-sample benchmark that tells them which short-read method to trust.
Terzian and colleagues built that benchmark in pig (Sus scrofa) and Japanese quail (Coturnix japonica). From a single blood DNA extraction per species, they generated parallel CpG methylation datasets on four sequencing platforms — including EM-seq libraries built from 200 ng of DNA fragmented on the PIXUL sonicator. EM-seq returned the highest mapping rates, the lowest duplicate rates, and the strongest agreement with long-read methylation calls of any short-read method tested.
The takeaway: a single PIXUL fragmentation step, paired with low-input enzymatic library prep, produces methylation data that holds up against three competing sequencing platforms.
Key findings
- EM-seq libraries built from 200 ng of pig and quail blood DNA — fragmented on the PIXUL sonicator ahead of NEBNext enzymatic conversion and 5 PCR cycles — produced sequencing-grade libraries that mapped at higher percentages than the matched WGBS controls built from the same extractions.
- EM-seq duplicate rates fell around 10%, well below the matched WGBS controls, and WGBS covered only 76.4% of reference CpG sites in quail versus near-complete site coverage from EM-seq and long reads — consistent with enzymatic conversion preserving DNA integrity better than bisulfite chemistry on PIXUL-fragmented input.
- Pearson correlations between EM-seq and Oxford Nanopore methylation calls reached 0.93 in quail and 0.92–0.97 in pig — the strongest short-read-to-long-read agreement in a four-platform comparison — positioning the PIXUL-prepped EM-seq pipeline as a reliable short-read reference for benchmarking long-read methylation callers.
PIXUL in the methods
“EM-seq libraries were prepared following NEB’s protocol using the NEBNext Enzymatic Methyl-seq Kit. Briefly, 200 ng of DNA were fragmented by sonication using PIXUL sonicator (Active Motif), followed by 5 PCR cycles.”
— Terzian et al., Scientific Data (2025), Methods, Library preparation and sequencing — Enzymatic Methyl-seq
Why it matters for PIXUL users
If you build sequencing libraries from genomic DNA, this paper is direct peer-reviewed evidence that PIXUL delivers the upstream fragmentation step inside a published EM-seq workflow benchmarked against three competing sequencing platforms. The authors fragmented 200 ng of pig and Japanese quail blood DNA on PIXUL ahead of NEB enzymatic library prep, and the resulting libraries returned higher mapping rates, lower duplicate rates (~10%), and the strongest Pearson agreement (0.93 in quail, 0.92–0.97 in pig) against Oxford Nanopore long-read methylation calls of any short-read method tested. For your own work in livestock genomics, comparative DNA methylation profiling, or low-input sequencing library prep, the practical takeaway is that PIXUL’s parallel sample preparation format integrates cleanly into low-input enzymatic workflows, that the fragmented DNA performs well enough to anchor a four-platform CpG methylation reference dataset, and that one instrument supports both your sequencing library prep and the protein and multi-omic workflows already documented elsewhere in the PIXUL publication record.
