Impact of sequence variant detection and bacterial DNA extraction methods on the measurement of microbial community composition in human stool

Abstract

Background: The human gut microbiome has been widely studied in the context of human health and metabolism, however the question of how to analyze this community remains contentious. This study compares new and previously well established methods aimed at reducing bias in bioinformatics analysis (QIIME 1 and DADA2) and bacterial DNA extraction of human fecal samples in 16S rRNA marker gene surveys. Results: Analysis of a mock DNA community using DADA2 identified more chimeras (QIIME 1: 0.70% of total reads vs DADA2: 1.96%), fewer sequence variants, (QIIME 1: 1297.4 �� 98.88 vs. DADA2: 136.27 �� 11.35, mean �� SD) and correct taxa at a higher resolution of classification (i.e. genus-level) than open reference OTU picking in QIIME 1. Additionally, the extraction of whole cell mock community bacterial DNA using four commercially available kits resulted in varying DNA yield, quality and bacterial community composition. Of the four kits compared, ZymoBIOMICS DNA Miniprep Kit provided the greatest yield, with a slight enrichment of Enterococcus. However, QIAamp Fast DNA Stool Mini Kit resulted in the highest DNA quality. Mo Bio PowerFecal DNA Kit had the most dramatic effect on the mock community composition, resulting in an increased proportion of members of the family Enterobacteriaceae and genus Eshcerichia as well as members of genera Lactobacillus and Pseudomonas. The presence of a sterile fecal matrix had a slight, but inconsistent effect on the yield, quality and taxa identified after extraction with all four DNA extraction kits. Extraction of bacterial DNA from native stool samples revealed a distinct effect of the DNA stabilization reagent DNA/RNA Shield on community composition, causing an increase in the detected abundance of members of orders Bifidobacteriales, Bacteroidales, Turicibacterales, Clostridiales and Enterobacteriales. Conclusion: These results confirm that the DADA2 algorithm is superior to sequence clustering by similarity to determine microbial community structure. Additionally, commercially available kits used for bacterial DNA extraction from fecal samples have some effect on the proportion of high abundance members detected in a microbial community, but it is less significant than the effect of using DNA stabilization reagent, DNA/RNA Shield.