Related Experiment Video
Updated: Dec 30, 2025
11:22
Microbiota Analysis Using Two-step PCR and Next-generation 16S rRNA Gene Sequencing
Published on: October 15, 2019
30.9K
Comparing bioinformatic pipelines for microbial 16S rRNA amplicon sequencing.
Andrei Prodan1, Valentina Tremaroli2, Harald Brolin2
1Department of Experimental Vascular Medicine, Amsterdam University Medical Centers, Amsterdam, The Netherlands.
Plos One
|January 17, 2020
Summary
This study compared six bioinformatic pipelines for microbial amplicon sequencing data analysis. USEARCH-UNOISE3 and DADA2 showed strong performance, offering guidance for accurate bacterial community analysis.
Area of Science:
- Microbiology
- Bioinformatics
- Genomics
Background:
- Microbial amplicon sequencing, particularly 16S rRNA gene surveys, is crucial for understanding bacterial ecosystems, including the human microbiome.
- Converting raw sequencing data into meaningful bacterial counts requires specialized bioinformatic pipelines.
- Current pipelines have limitations and biases, leading to a lack of standardized best practices.
Purpose of the Study:
- To compare the performance of six different bioinformatic pipelines for microbial amplicon sequence data analysis.
- To evaluate pipelines at both Operational Taxonomic Unit (OTU) and Amplicon Sequence Variant (ASV) levels.
- To provide guidance on pipeline selection for accurate microbial community profiling.
Main Methods:
- Six pipelines were tested: QIIME-uclust, MOTHUR, USEARCH-UPARSE (OTU-level), and DADA2, Qiime2-Deblur, USEARCH-UNOISE3 (ASV-level).
- Analyses included varying quality control, clustering algorithms, and cutoff parameters.
- Performance was assessed using a mock community and a large fecal sample dataset (N=2170) from the HELIUS study.
Main Results:
- DADA2 demonstrated the highest sensitivity but lower specificity compared to USEARCH-UNOISE3 and Qiime2-Deblur.
- USEARCH-UNOISE3 offered an optimal balance between resolution and specificity.
- OTU-level pipelines (USEARCH-UPARSE, MOTHUR) performed well but had lower specificity than ASV-level methods.
- QIIME-uclust generated numerous spurious OTUs and inflated diversity measures, recommending its avoidance.
Conclusions:
- ASV-level pipelines, particularly USEARCH-UNOISE3, generally provide superior accuracy for microbial amplicon data analysis.
- Pipeline choice significantly impacts results, necessitating careful consideration of sensitivity, specificity, and resolution.
- This comparative study offers valuable insights for researchers to optimize their bioinformatic workflows for microbial ecology studies.
