In gut microbial research, DNA extraction remarkably influences study outcomes and biological interpretations. Rapid advancements in the research scale and technological upgrades necessitate evaluating new methods to ensure reliability and precision in microbial community profiling. We systematically evaluated the performance of eight recent and commonly used extraction methods using a microbial mock community (MMC) and fecal samples from two healthy volunteers, incorporating bacterial, archaeal, and fungal constituents. Performance metrics included nucleic acid assessment, microbial profile assessment, and scalability for large-scale studies, leveraging shotgun metagenomics for in-depth analysis. Despite variations in DNA quantity and quality, all methods yielded sufficient DNA for shotgun metagenomic sequencing. In the MMC microbial profile assessment, the QIAamp PowerFecal pro Kit (PF) and DNeasy PowerSoil HTP kit (PS) methods exhibited higher similarity with the theoretical composition and lower variability across technical replicates compared to other methods. For fecal samples, the extraction method accounted for 21.4% of the overall microbiome variation and significantly affected the abundances of 32% of detected microbial species. Methods using mechanical lysis with small beads, such as PF and PS, demonstrated better efficiency, indicated by increased microbial diversity in extracting DNA from Gram-positive bacteria. Furthermore, the PF and PS methods are notably simple to execute and automation-friendly, though relatively costly. Our study underscores the importance of maintaining consistency in DNA extraction methods for reliable comparative metagenomic analyses. We recommend PF and PS methods as optimal for expansive gut metagenomic research, emphasizing the critical role of mechanical lysis in DNA extraction.
Supplementary information: The online version contains supplementary material available at 10.1007/s43657-025-00232-x.
Keywords: DNA extraction; Gut microbiota; Method evaluation; Shotgun metagenomics.
© International Human Phenome Institutes (Shanghai) 2025. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.