Related Experiment Video
Updated: May 17, 2026
11:23
Purifying the Impure: Sequencing Metagenomes and Metatranscriptomes from Complex Animal-associated Samples
Published on: December 22, 2014
Denoising PCR-amplified metagenome data
Michael J Rosen1, Benjamin J Callahan, Daniel S Fisher
1Department of Applied Physics, Stanford University, CA, USA. mjrosen@stanford.edu
BMC Bioinformatics
|November 2, 2012
Summary
A new Divisive Amplicon Denoising Algorithm (DADA) accurately identifies microbial and viral diversity by inferring genotypes and error parameters. This fast, accurate method distinguishes true biological variation from sequencing errors without training data.
Area of Science:
- Microbial ecology
- Metagenomics
- Bioinformatics
Background:
- High-throughput sequencing and PCR amplification generate errors that obscure true microbial and viral diversity.
- Existing denoising methods often lack sufficient speed or accuracy for comprehensive analysis.
Purpose of the Study:
- To introduce a novel denoising algorithm, Divisive Amplicon Denoising Algorithm (DADA), for accurate metagenomic data analysis.
- To develop a method that can infer sample genotypes and error parameters without requiring training data.
Main Methods:
- DADA algorithm infers genotypes and error parameters directly from metagenomic data.
- Performance evaluated on Roche 454 platform data.
- Comparison with existing denoising software, AmpliconNoise.
Main Results:
- DADA demonstrates superior accuracy compared to AmpliconNoise.
- DADA is over an order of magnitude faster than AmpliconNoise.
- The algorithm effectively utilizes sequence-abundance information and incorporates context-dependent PCR error rates.
Conclusions:
- DADA provides a significant advancement in denoising metagenomic data, offering both speed and accuracy.
- The method eliminates the need for training data, simplifying the analysis pipeline.
- DADA is adaptable for other sequencing platforms like Illumina, broadening its applicability.
