Impact of Fecal Microbiome Extraction Technique on Relative Abundance of Genera within Expected and Unexpected Communities.

Microbiome analysis has gained significant interest as sequencing technology has improved, allowing for the examination of microbial communities often believed to play a key role in human health and disease. 16S rRNA sequencing analysis, utilizing various publicly available bioinformatic tools, allows the study of diversity in microbial communities. It has previously been shown that extraction techniques are impactful for the analysis, however, these studies have often been performed with a single cultured sample, or with evenly distributed mock communities, failing to provide variability often seen with human source samples and excluding the impact of amplification and/or sequencing related bias. Within this study, we analyzed six different extraction protocols, utilizing commercially available reagents, with a naturally collected fecal sample, two mock communities, all of uneven species distribution, and a manufactured extraction control, of known species distribution. In addition, six sequencing controls, already extracted DNA of variable concentration and percentage of species distributed within each community, were utilized to determine the additional bias introduced by amplification and/or sequencing reagents. Significant differences were noted in alpha and beta diversity of samples extracted using each protocol, particularly as compared to the expected values, consistent with previous findings. Additionally, there were unexpected genera found within these communities indicating contamination uniquely introduced by each protocol. Trends were noted among different kits, in either over- or under-representation of specific genera of communities. Similar trends were observed in the results of sequencing controls, indicating the contamination and/or bias introduced after extraction. Overall, these findings illustrate the importance of utilizing controls at multiple stages of the process, to better assess the impact of the findings of any human samples.