Metagenomic Negative Controls

Metagenomic negative controls are samples processed alongside your biological specimens that contain no intended template DNA. They capture contamination introduced during collection, extraction, library preparation, and sequencing. Without them, signals from contaminants can be mistaken for true microbial diversity, especially in low biomass environments. This guide is for researchers and bioinformaticians designing metagenomic studies who need a practical, source bounded framework for implementing negative controls correctly. The Galaxy Training Network offers workflow materials that complement these concepts. You will learn core concepts, decision points, a step by step workflow, quality checks, common mistakes, and limits of interpretation. The EMBL EBI Training resources provide additional background on sequencing best practices.

At a Glance

Core Concepts of Metagenomic Negative Controls

A negative control is a sample that undergoes the same processing steps as your metagenomic samples but contains no deliberately added microbial DNA. Its sole purpose is to capture the background signal introduced by your workflow. The NCBI Bookshelf contains authoritative technical references on contamination controls in sequencing. There are several types of negative controls used in metagenomics:

• Kit blank: A control that goes through the DNA extraction kit using only the kit reagents, without any sample.
• Extraction blank: A tube with sterile water or buffer that is processed through the entire extraction procedure.
• PCR blank: A no template control that goes only through the PCR amplification step.
• Water control: A sample consisting solely of molecular grade water sequenced to assess lab water quality.
• Sterile swab: For studies using swabs, a swab exposed to air or the collection environment.

All these controls help identify the “kitome” (contamination from kit reagents) and the “labome” (contamination from lab air and surfaces). A study on betel quid chewers used negative controls to rule out reagent contamination in microbiome gut brain axis analysis [6]. In clinical metagenomics, controls are essential for distinguishing true pathogens from background flora [7].

Decision Points for Choosing Negative Controls

Not all negative controls are equally informative for every study. You must choose control types based on your sample type, biomass level, and research question. Consider these criteria:

• Sample biomass: Low biomass samples (cerebrospinal fluid, bronchoalveolar lavage, tissue biopsies) are highly susceptible to reagent contamination. Include both kit blanks and extraction blanks. For high biomass samples like feces, a single water control may suffice.
• Environmental exposure: If you collect samples in a hospital or farm environment, include a swab that samples the air or surface at the collection site. A study on dental unit waterlines used environmental controls to assess background contamination [11].
• Sequencing depth: Sequence your negative controls to at least 10% of the depth of your biological samples. Low depth controls may miss low level contamination. The NCBI Sequence Read Archive provides examples of control sequencing depths in public datasets.
• Number of controls: Include at least one control per 20 biological samples, and always include one control per batch. Multiple controls help distinguish steady contamination from sporadic events.
• Budget constraints: If resources are limited, prioritize extraction blanks and PCR blanks over kit blanks, as these capture more of the workflow.

Practical Workflow for Implementing Negative Controls

Follow this sequence to integrate negative controls into your metagenomic experiment:

1. Plan control inclusion during experimental design. Decide on control types and numbers based on the decision criteria above.
2. Collect controls at each stage:
   • At extraction: prepare one extraction blank for every batch of 20 samples. Use the same reagents and tubes.
   • At PCR: include a PCR blank (no template DNA) in every amplification plate.
   • At library preparation: include a library blank if you add indexing steps.
3. Process controls in parallel with biological samples. Do not assign them later, they must undergo identical handling.
4. Sequence the controls: Pool controls into the same run as your samples. Use different barcodes to identify them. The Bioconductor project offers packages like metagenomeSeq that can analyze control contamination.
5. Bioinformatics processing:
   • Remove adapter sequences and low quality reads from all samples, including controls.
   • Map reads to common contaminant databases (human, PhiX, known reagent contaminants).
   • Generate taxonomic profiles for controls using the same pipeline as your samples.
   • Subtract contaminant features from biological samples. Tools like decontam in R (available via Bioconductor) use prevalence or frequency to identify contaminants.
6. Document all control data: Record batch numbers of all reagents, control read counts, and taxonomic assignments. This aids troubleshooting later.

The Galaxy Training Network has tutorials for metagenomic analysis that include steps for handling negative controls.

Quality Checks for Negative Control Data

Once you have sequenced and analyzed your negative controls, evaluate them with these checks:

• Check read counts: Controls should have far fewer reads than biological samples. If a control has more reads than expected, investigate potential index hopping or mislabeling. A study on probiotics in fish included control checks to validate gut microbiome results [8].
• Assess taxonomic diversity: A typical negative control will have a few contaminants: common environmental bacteria (e.g., Ralstonia, Pseudomonas), skin flora (Staphylococcus, Corynebacterium), and lab reagents (E. coli from cloning strains). If you see unusual or abundant taxa, that indicates a contamination event.
• Correlate with sample profiles: Compare the most abundant taxa in controls to those in biological samples. If a taxon is abundant in both, it might be a contaminant rather than a true signal. Use correlation analysis and prevalence across samples.
• Use quality thresholds: A common threshold is that any taxon present in a negative control at greater than 1% relative abundance should be removed from all samples unless there is evidence it is endogenous (e.g., expected from the sample site). The journal recommendations from NCBI Bookshelf support such thresholds.
• Repeat controls: In a new batch of sequencing, include duplicate controls to assess reproducibility of contamination patterns. This is especially important in clinical metagenomics where results guide treatment decisions [10].

Common Mistakes with Negative Controls

• Omitting controls entirely: Many early metagenomic studies lacked negative controls, leading to published claims that were later attributed to contaminants. Always include them.
• Using only one type of control: A single water control does not capture contamination from extraction columns or PCR enzymes. Use multiple types.
• Sequencing controls too shallow: If a control gets only 100 reads, you cannot detect low level contamination that might still affect low biomass samples. Sequence controls to at least 50,000 reads for shotgun metagenomics.
• Processing controls in a separate session: If you handle controls in a different lab or on a different day, they do not represent the actual contamination risk to your samples. Process everything together.
• Ignoring low level contamination: Just because a contaminant appears at 0.1% in a control does not mean it is negligible. In a low biomass sample with few reads, that 0.1% could become a dominant signal.
• Failing to include controls in bioinformatics pipelines: Do not just look at control data manually. Integrate it into your analysis with statistical decontamination methods available in Bioconductor packages. A study on antimicrobial resistance in respiratory microbiomes used controls to correctly attribute resistance genes to pathogens versus contaminants [9].

Limits of Interpretation for Negative Controls

Negative controls are powerful but they have inherent limits.

• They cannot capture all contamination sources. For example, cross contamination between samples during multiplexing (index hopping) is not captured by a separate control. Use unique dual indices and controls with known barcodes.
• Reagent batch variation. Contamination profiles can change with new lots of kits. A control from last month does not apply to today’s run. Always run controls with each reagent batch.
• Low biomass samples have very high sensitivity to contamination. Even with controls, some contaminant sequences may escape detection if they are present only in a subset of samples. The EMBL EBI Training materials emphasize the statistical challenges of low biomass metagenomics.
• Subtraction may remove true signals. If a microbe is truly present at low abundance in your sample and also appears in the control (from a previous sample on the same sequencing run), subtraction could erase it. Use frequency based methods rather than simple subtraction to reduce this risk.
• Controls do not correct for amplification bias. PCR biases that favor certain templates affect controls and samples differently. Negative controls can underestimate contamination if the contaminant DNA has a high GC content that amplifies poorly with your polymerase.
• Interpretation requires biological context. A taxon in a control may be a true environmental microbe that also colonizes your sample site (e.g., skin). Statistical decontamination cannot always resolve this. The clinical impact study on 16S rRNA NGS [7] discusses how clinical correlation is needed.

Always interpret negative control results as one part of a quality control framework. Do not rely solely on automated subtraction. Manual inspection of control profiles and comparison with blank sites (e.g., from NCBI Sequence Read Archive public data) can improve your confidence.

Frequently Asked Questions

1. How many negative controls should I include in my metagenomic study?

Include at least one control per 20 biological samples per batch. For low biomass studies, include two controls per batch. Always include at least one extraction blank and one PCR blank. If you use multiple reagent lots, include a control for each lot.

2. Can I use nuclease free water as a negative control?

Yes, molecular grade nuclease free water is a standard control for water quality and PCR steps. But water controls do not capture contamination from extraction columns or bead beating tubes. Combine water controls with extraction blanks for full coverage.

3. What does it mean if my negative control has a lot of reads?

A negative control with many reads indicates significant contamination. Check whether the control was properly barcoded and not mislabeled. Common reasons: old reagents, inadequate lab hygiene, or index hopping. Do not proceed with sample analysis until you identify and remedy the source.

4. Should I subtract all sequences found in negative controls from my samples?

No. Simple subtraction removes reads that may be true low abundance members of your community. Instead, use statistical methods (e.g., using prevalence of a taxon across samples vs. its abundance in controls) to flag contaminants. Only remove features that are significantly enriched in controls compared to biological samples and that are known lab contaminants.

References and Further Reading

• NCBI Bookshelf , Free biomedical books with chapters on sequencing contamination.
• EMBL EBI Training , Official training on metagenomics quality control.
• Galaxy Training Network , Tutorials for metagenomic analysis including negative controls.
• Bioconductor , R packages for decontamination and analysis.
• NCBI Sequence Read Archive , Repository for public metagenomic data and control metadata.
• Study on betel quid chewers: microbiome gut brain axis , Example of negative control use in gut microbiome study.
• Clinical impact of 16S rRNA NGS , Paper discussing controls in clinical diagnostics.
• Probiotic study on fish gut microbiome , Example of negative controls in animal studies.
• Respiratory tract resistome study , Use of controls in antimicrobial resistance profiling.
• mNGS for elderly stroke patients , Clinical metagenomics with control integration.
• Dental unit waterline study , Environmental control application.

Related Articles

• Read Duplication Sequencing
• Per Base Quality Sequencing
• Kmer Spectrum Analysis
• Genome Size Estimation Kmer
• Contamination Screening Genome Assembly