High-resolution shotgun metagenomics: the more data, the better?

Area of Science:

• Microbial Ecology
• Bioinformatics
• Genomics

Background:

• High-resolution shotgun metagenomics (HRSM) workflows are computationally intensive and require significant storage.
• Increasing DNA sequencing output necessitates workflow adjustments for processing large datasets.
• Shallow shotgun metagenomics (SM) is a potential adaptation to manage growing data volumes.

Purpose of the Study:

• To benchmark the viability of shallow SM datasets using real-world data.
• To assess if reduced sequencing depth impacts microbial community structure analysis.
• To determine the optimal sequencing depth for ecological interpretations in SM.

Main Methods:

• Utilized four public SM datasets of varying sizes (one massive, three moderate).
• Subsampled datasets at multiple levels to simulate various shallow sequencing depths.
• Compared results from shallow datasets with those from full-depth datasets.

Main Results:

• Shallow SM sequencing effectively captures microbial community structures.
• Subsampling down to 0.5 million sequencing clusters per sample yielded comparable results to larger datasets for human gut and agricultural soil.
• For an Antarctic dataset, 4 million sequencing clusters per sample were sufficient for comparable results.
• Ultra-deep sequencing was beneficial primarily for generating metagenome-assembled genomes.

Conclusions:

• Shallow shotgun metagenomics is a viable and cost-effective approach for microbial community structure analysis.
• High-depth sequencing offers limited additional ecological interpretation benefits over optimized shallow sequencing.
• Shallow sequencing can significantly reduce computational and storage requirements for metagenomic studies.