DNA Reference Reagents Isolate Biases in Microbiome Profiling: A Global Multi-Lab Study

When profiling the human gut microbiome, technical biases introduced by analytical approaches impede translational research, reducing data reliability and study comparability. Here, through a global study involving 23 labs, we analysed a wide range of sequencing and bioinformatic approaches for the taxonomic profiling of two well-defined DNA reference reagents (RRs) comprised of 20 common gut bacteria. Through both shotgun and 16S rRNA gene amplicon sequencing we aimed to isolate sources of bias and understand their impact on microbiome profiling accuracy. Importantly, minimum quality criteria (MQC) were established and are used to evaluate profiling performance. We found variability of shotgun sequencing datasets was greater than that of 16S rRNA gene amplicon sequencing and isolated sources of bias in wet and dry lab steps, such as sequencing depth, primer and database choices, rarefaction and 16S copy number adjustment. This study presents well-defined RRs and MQC to combat technical bias, paving the way for reliable and comparable microbiome research.