Quantifying and understanding well-to-well contamination in microbiome research

When Amnon Amir, Jon Sanders, and their colleagues began using positive control samples of Vibrio fischeri in plate-based extractions as a way to validate results, they unexpectedly observed that many of the surrounding samples would also show Vibrio fischeri in their composition. They set out to design an experiment to quantify this problem, which led to the newly published paper "Quantifying and Understanding Well-to-Well Contamination in Microbiome Research," in mSystems.

"We were quite surprised by both the span of well-to-well contamination (number of samples affected) along with the total percentage of reads from a given sample being derived from well-to-well contamination," states Jeremiah Minich, lead author on the paper. They discovered that well-to-well contamination is a major source of contamination in microbiome studies, particularly impacting low biomass samples the most. Minich and his team found that the DNA extraction method is the primary driver of well-to-well contamination.

In short, if one were to do robot-based extractions, particularly when doing the upfront lysis protocol, one would run the risk of a much higher rate (~2x) of contamination than if someone used single tube extractions.

"Any existing pipeline developed for predicting disease outcomes or identifying clinically relevant microbes associated with disease state will need to consider and estimate the noise associated with well-to-well contamination in a study design to improve diagnosis accuracy and resolution. Specifically, clinical microbiome methods which simply subtract out reads or microbes found in 'negative' samples will likely be excluding 'real' microbes found in other samples," explains Minich on how this may impact studies of the future.

He believes companies will need to figure out and design kits which do not have so much well-to-well contamination and should consider reporting these levels on the label. For now, they have developed and implemented a hybrid approach to reduce well-to-well contamination which includes a lysis steps in the single tube followed by magnetic bead cleanup which is believed to reduce well-to-well contamination while also enabling low biomass detection.