Can the fragmentation enzyme in the library preparation kit replace sonication shearing?

Many enzyme fragmentation + library preparation kits can indeed be compressed to about half a day, but this does not cover the entire process from sample to completion of QC before sequencing. Generally, fragmentation, end repair, A-tailing, and adapter ligation can be completed within 2.5-3 hours. Adding magnetic bead purification, PCR, and library QC, the actual total time is about 6-8 hours. If the sample volume is large, using 96-well plates and multichannel pipettes, starting in the morning, purification can be completed and the library submitted before the end of the workday in the afternoon. However, the enzyme digestion time must be optimized according to the target insert size and cannot be fixed, otherwise the fragments will be too small. It is recommended to first perform a gradient pilot experiment, and then compress the formal library preparation into a half-day workflow.

If your main goal is to shorten the fragmentation time, you can look at kits such as KAPA HyperPlus and NEB Ultra II FS, which combine fragmentation, end repair, and A-tailing, and domestic brands such as Vazyme and Diying also have similar types. Metagenomic DNA input is usually not low, so the amount of fragmentation enzyme can be fine-tuned according to the template amount. For high-GC samples, I would set the fragmentation time shorter and run preliminary experiments at several time points. Dual-round magnetic bead selection after fragmentation is also critical, as it can remove small fragments and avoid wasting sequencing data. The total duration can be kept within half a day, and operationally it is recommended to equilibrate the magnetic beads to room temperature in advance and preheat the PCR instrument, which can save a lot of waiting.

Enzymatic fragmentation can replace sonication in routine DNA library preparation, especially when sample throughput is high, as it eliminates the separate steps of fragmentation and end repair. However, for bacterial metagenomics, I would not recommend direct substitution. In metagenomic samples, GC content varies widely across species, and genome sizes differ; enzymatic fragmentation bias may lead to lower coverage in certain high-GC or low-GC regions, causing bias in downstream binning or assembly. We once performed a parallel comparison on the same batch of samples: the enzymatic protocol produced a more concentrated fragment size distribution, but coverage uniformity was slightly worse than sonication. You can first run a small-scale comparison of sonication vs. enzymatic fragmentation, evaluating GC bias and taxonomic annotation, before deciding whether to scale up to large samples.

Fragmentation enzymes are usually not completely random. They often have enzyme recognition sites or transposase insertion biases. If high-GC regions form secondary structures, digestion efficiency may decrease, leading to systematic undercoverage of these regions. In metagenomics, high-GC actinobacteria and some archaea are more susceptible. It is recommended not to rely solely on manufacturers' claims of "random fragmentation." You can spike known high-GC positive bacteria into mock samples, then perform library preparation and sequencing using sonication and enzymatic digestion separately, and compare relative abundance and coverage depth. If obvious bias occurs, it is better to retain sonication, even if using a multi-channel sonication device; alternatively, extend the enzymatic digestion time and increase the temperature, but this may introduce more small fragments.

If your downstream focus is metagenomic assembly, MAG binning, or SV analysis, I would not recommend long-term use of enzymatic fragmentation as a replacement for sonication. Enzymatic bias can reduce the continuity of complex community assemblies, and low-abundance species in particular may lose fragments due to the distribution of cutting sites. When sample numbers are large, sonication is time-consuming, but conditions can be optimized: first quantify and fix the sample volume, use a multi-channel sonication probe to interrupt samples in batches, or introduce an automated workstation for subsequent library preparation. If enzymatic fragmentation must be used, the QC standards for each project can be raised, for example by monitoring the proportion of low-coverage regions and the GC bias curve, and reverting to sonication if abnormalities appear.