The human gut microbiome is receiving increasingly more attention in recent years due to growing awareness of its relationship to human health. Taxonomic identification and phylogenetic profiling are the fundamental steps in the characterization of the gut microbiome. However, many recent publications have highlighted data variations of gut microbiome profiling when utilizing different methods for sample preservation, DNA purification, library preparation, and bioinformatics analysis. In this study, we illustrate key biases and variations caused by improper sample processing and present optimized solutions to prevent such discrepancies. We found liquid preservatives protect and preserve the composition of a microbiome sample prior to DNA extraction. Also, exaggerated variations of microbial composition can stem from the use of different DNA extraction kits. Additionally, differences in NGS library preparations can have a dramatic effect on the final composition profile. And finally, bioinformatic analysis can have a large impact on the observed microbial profiles. The microbiome profiling workflow has many sequential steps where biases can compound, amplifying deviations from the true profile, and the complexity of fecal samples further convolutes this process. But, if these sources of bias are considered and validated methods are used, researchers can have confidence in the accuracy of microbiome data, from which meaningful conclusions can be drawn.
展开▼
