Multicenter quality assessment of 16S ribosomal DNA-sequencing for microbiome analyses reveals high inter-center variability

Abstract

The composition of human as well as animal microbiota has increasingly gained in interest since metabolites and structural components of endogenous microorganisms fundamentally influence all aspects of host physiology. Since many of the bacteria are still unculturable, molecular techniques such as high-throughput sequencing have dramatically increased our knowledge of microbial communities. The majority of microbiome studies published thus far are based on bacterial 16S ribosomal RNA (rRNA) gene sequencing, so that they can, at least in principle, be compared to determine the role of the microbiome composition for host metabolism and physiology, developmental processes, as well as different diseases. However, differences in DNA preparation and purification, 16S rDNA PCR amplification, sequencing procedures and platforms, as well as bioinformatic analysis and quality control measures may strongly affect the microbiome composition results obtained in different laboratories. To systematically evaluate the comparability of results and identify the most influential methodological factors affecting these differences, identical human stool sample replicates spiked with quantified marker bacteria, and their subsequent DNA sequences were analyzed by nine different centers in an external quality assessment (EQA). While high intra-center reproducibility was observed in repetitive tests, significant inter-center differences of reported microbiota composition were obtained. All steps of the complex analysis workflow significantly influenced microbiome profiles, but the magnitude of variation caused by PCR primers for 16S rDNA amplification was clearly the largest. In order to advance microbiome research to a more standardized and routine medical diagnostic procedure, it is essential to establish uniform standard operating procedures throughout laboratories and to initiate regular proficiency testing.