Fig. 1: Schematic overview of the work.

a The use of reference communities of microbes in vitro and in vivo using data from the literature or generated in-house and analyzed using different bioinformatic pipelines allowed precise analysis of the occurrence of spurious taxa. b Additional experiments using cecal contents from germfree mice in combination with DNase pre-treatment and mock DNA spiking were performed in a second sequencing lab to test effects of the matrix background on analysis outcomes. c Several human fecal samples stored under different conditions and processed in triplicates in different sequencing runs allowed assessing the reproducibility of microbiota profiles generated by high-throughput 16S rRNA gene amplicon sequencing following different filtering thresholds to remove spurious taxa. d Sequencing of a mock community and a soil sample, including several replicates and sequencing runs, followed by data analysis in a third facility were performed to validate findings. All technical details are given in the Methods section.