Table 1 Guidelines for minimizing contamination when conducting marker gene or metagenomic sequencing-based studies of microbiomes
From: Guidelines for preventing and reporting contamination in low-biomass microbiome studies
Area | Approaches | Importance |
|---|---|---|
Awareness and training | Ensure continual awareness and care regarding the multiple sources and steps that can introduce contamination and cross-contamination, as well as the importance of including controls | Essential |
Prioritize contamination-aware approaches for every aspect of microbiome research, including laboratory setup, study design, laboratory procedures and data analysis, focusing on preventative rather than corrective approaches | Very high | |
Develop SOPs for all field, laboratory and computational procedures to minimize and detect contamination | Very high | |
Conduct both theoretical and hands-on training for field, laboratory and computational procedures to minimize and detect contamination | High | |
Provide regular reminders and conduct audits of protocols and practices to reduce contamination risk | Moderate | |
Develop continuous improvement processes where SOPs and training are updated in response to day-to-day practice and audits | High | |
Sample collection | Decontaminate sampling equipment and reagents to remove cells and any non-target DNA | High |
Design sampling approach to minimize introduction of DNA from environments adjacent to the environment of interest | High | |
If possible, collect samples from environments near the environment of interest that may be the source of potential contaminants | Moderate | |
Wear appropriate PPE and carefully handle samples to minimize introduction of DNA from human operators and their microbiomes | High | |
Collect control samples at all sample collection and processing steps to check for the introduction of potential contaminants, and include these control samples in all downstream processing steps | Very High | |
Store samples in a manner that eliminates the potential for microbial growth or contamination with foreign DNA | High | |
Laboratory practices | Include appropriate negative controls at all processing steps and include all negative controls in the sequencing runs | Essential |
Include positive controls, ideally at microbial DNA concentrations that mirror those in samples | High | |
Determine the concentrations of DNA in negative controls via qPCR to assess signal-to-noise ratios compared with samples | High | |
Conduct test runs to confirm suitability of DNA extraction, library preparation and sequencing procedures before conducting full sampling campaigns | High | |
Conduct molecular work in hoods to reduce contamination with laboratory operators, materials and air | Very high | |
Use a unidirectional workflow, including separating workspaces and equipment for pre- and post-extraction and pre- and post-PCR procedures | High | |
Thoroughly decontaminate hoods/workspaces, equipment and consumables to both kill cells and remove DNA | Very high | |
Use trusted reagent suppliers and sequencing facilities | High | |
Test each individual batch of reagents for potential contaminants before use | High | |
Wear appropriate PPE to minimize the introduction of human-derived contaminants | Very high | |
Take utmost care to minimize sample-to-sample cross-contamination, and avoid processing higher- and lower-biomass samples together | High | |
Consider the spatial arrangement of samples and controls in multiwell plates during extraction, library preparation and sequencing to reduce cross-contamination and ‘batch effects’ | High | |
Uniquely barcode samples and controls before sequencing | High | |
Data analysis and reporting | Follow ‘minimal reporting standards’ (Table 2) to ensure sampling, laboratory and analysis procedures are transparently documented in full, and both raw and processed data are reported and accessible | Essential |
Carefully check sequence data from negative and positive controls to infer the presence, extent and potential sources of any contamination or cross-contamination | Very high | |
Check whether the major taxa observed in samples are expected on the basis of a priori knowledge of their potential ecologies and known contaminants | Moderate | |
Consider mapping to reference genomes of potential contaminants, for example, human genome, human-associated microbes and reagent microbes | Moderate | |
Consider identifying potential contaminants using decontamination bioinformatics tools, but do so carefully, critically and with transparency | High | |
When possible, collect sequence data along with independent measures of microbial DNA concentrations (or microbial biomass), as contaminants will generally be enriched in samples with lower microbial DNA concentrations | Moderate | |
Only use contaminated samples or datasets if sufficient signal to noise can be recovered to convincingly address research questions | Very high | |
Report the concentrations of 16S rRNA genes in negative controls alongside samples to assess signal to noise | High |
