Multicenter assessment of microbial community profiling using 16S rRNA gene sequencing and shotgun metagenomic sequencing

• Published in: Journal of advanced research Vol. 26; pp. 111 - 121
• Main Authors: Han, Dongsheng, Gao, Peng, Li, Rui, Tan, Ping, Xie, Jiehong, Zhang, Rui, Li, Jinming
• Format: Journal Article
• Language: English
• Published: Egypt Elsevier B.V 01.11.2020; Elsevier
• Subjects: 16S rRNA gene sequencing; Microbial community profiling; Microbiome; Microbiota; Shotgun metagenomic sequencing; Shotgun metagenomic sequencing; Microbiota; 16S rRNA gene sequencing; Microbial community profiling; Microbiome
• DOI: 10.1016/j.jare.2020.07.010
• ISSN: 2090-1232

[Display omitted] Microbiome research based on high-throughput sequencing has grown exponentially in recent years, but methodological variations can easily undermine the reproducibility across studies. To systematically evaluate the comparability of sequencing results of 16S rRNA gene sequencing (16Ss)- and shotgun metagenomic sequencing (SMs)-based microbial community profiling in laboratories under routine conditions. We designed a multicenter study across 35 participating laboratories in China using designed mock communities and homogenized fecal samples. A wide range of practices and approaches was reported by the participating laboratories. The observed microbial compositions of the mock communities in 46.2% (12/26) of the 16Ss and 82.6% (19/23) of the SMs laboratories had significant correlations with the expected result (Spearman r>0.59, P <0.05). The results from laboratories with near-identical protocols showed slight interlaboratory deviations. However, a high degree of interlaboratory deviation was found in the observed abundances of specific taxa, such as Bacteroides spp. (range: 0.3%-53.5%), Enterococci spp. (range: 0.8%-43.9%) and Fusobacterium spp. (range: 0.1%-39.8%). SMs performed better than 16Ss in detecting low-abundance bacteria (B. bifidum). The differences in DNA extraction methods, amplified regions and bioinformatics analysis tools (taxonomic classifiers and database) were important factors causing interlaboratory deviations. Addressing laboratory contamination is an urgent task because various sources of unexpected microbes were found in negative control samples. Well-defined control samples, such as the mock communities in this study, should be routinely used in microbiome research for monitoring potential biases. The findings in this study will provide guidance in the choice of more reasonable operating procedures to minimize potential methodological biases in revealing human microbiota composition.