# Bioconductor Dada2

> The dada2 package infers exact amplicon sequence variants (ASVs) from high-throughput amplicon sequencing data, replacing the coarser and less accurate OTU clustering approach. The dada2 pipeline takes as input demultiplexed fastq files, an

- Skill: `biomate-ai/bioconductor-dada2` (Agent Skill)
- Install (CLI): `npx skillmds@latest add biomate-ai/bioconductor-dada2`
- Raw SKILL.md: https://api.skillmd.com/api/skills/biomate-ai/bioconductor-dada2/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: DevOps & Infra
- Author: biomate-ai (https://skillmd.com/u/biomate-ai)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/biomate-ai/bioconductor-dada2

---


# dada2

## Dependencies & Environment

> Package-intrinsic requirements from the Bioconductor landing page — reproduce in any R environment.

- **Version:** 1.40.0 · **Bioconductor:** 3.23 · **R:** ≥ 4.6
- **Depends:** Rcpp
- **Imports:** Biostrings, ggplot2, reshape2, ShortRead, RcppParallel, IRanges, XVector, BiocGenerics
- **System requirements:** GNU make
- **Install:** `BiocManager::install("dada2")`

## When to Use
- Resolving exact Amplicon Sequence Variants (ASVs) from high-throughput amplicon sequencing data using the `dada` algorithm.
- Filtering and trimming raw paired-end or single-end FASTQ files using `filterAndTrim`.
- Modeling and learning sequencing error rates directly from the data using `learnErrors`.
- Removing chimeric sequences from merged sequence tables using `removeBimeraDenovo`.

## When NOT to Use
- For shotgun metagenomic sequencing data (use MetaPhlAn or HUMAnN instead).
- For clustering sequences based on a fixed 97% identity threshold (use vsearch or UCLUST instead).
- For downstream ecological statistics and visualization of ASV tables (use `phyloseq` instead).

## Data Requirements
- **Input Format**: Demultiplexed FASTQ files (unzipped or gzipped). One file per sample for single-end, or two files per sample for paired-end.
- **Quality Scores**: Raw reads must contain quality scores.
- **Primers**: Sequences must have non-biological bases (like PCR primers) removed prior to ASV inference.

## Key Parameters
- **`truncLen`**: Truncate reads after this many bases. Reads shorter than this are discarded. Crucial for removing low-quality tails.
- **`maxEE`**: After truncation, reads with higher than this number of expected errors are discarded (e.g., `maxEE=2`).
- **`trimLeft`**: The number of nucleotides to remove from the start of each read (useful for removing primers).
- **`maxN`**: Maximum number of ambiguous nucleotides allowed. DADA2 requires `maxN=0`.
- **`multithread`**: If TRUE, enables parallel processing to speed up computation.
- **`err`**: The error model generated by `learnErrors`, passed into the `dada` function.

## Best Practices
- Inspect the quality profiles of forward and reverse reads using `plotQualityProfile` before setting `truncLen` parameters.
- Learn error rates using `learnErrors` on at least a subset of your data to ensure accurate sample inference.
- Maintain a suitable overlap (>20nts) between forward and reverse reads after trimming to ensure `mergePairs` succeeds.
- Perform chimera removal using `removeBimeraDenovo` on the combined sequence table created by `makeSequenceTable`.

## Common Pitfalls
- **Failing to merge paired reads**: Truncating reverse reads too aggressively can prevent them from overlapping with forward reads, causing `mergePairs` to fail. Fix this by ensuring `truncLen` leaves enough sequence for a >20nt overlap.
- **Leaving primers in the sequences**: Non-biological nucleotides distort the error model. Fix this by using `trimLeft` in `filterAndTrim` or an external tool to remove primers.
- **Reads dropped due to Ns**: DADA2 cannot process sequences with ambiguous bases. Fix this by ensuring `maxN=0` is set in `filterAndTrim`.

## Alternatives
- **phyloseq**: Used for downstream analysis, visualization, and integration of the ASV tables produced by DADA2.
- **vsearch**: An alternative for traditional OTU clustering based on sequence identity thresholds.
- **deblur**: An alternative sub-operational taxonomic unit (sOTU) algorithm that uses a static error model.

## Citations
- Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJ, Holmes SP. (2016). "DADA2: High-resolution sample inference from amplicon data." *Nature Methods*, 13(7), 581-583.
- Callahan BJ, McMurdie PJ, Holmes SP. (2017). "Exact sequence variants should replace operational taxonomic units in marker-gene data analysis." *ISME Journal*, 11(12), 2639-2643.

## References
- Homepage: https://bioconductor.org/packages/dada2
- Vignette: https://bioconductor.org/packages/release/bioc/vignettes/dada2/inst/doc/dada2-intro.html

<!-- biomate-cta -->
---

## Run this on BioMate

This skill is the **knowledge layer** — when, why, and how to use `dada2`. To **run this analysis on your own data** with managed compute, automated QC, and reproducible outputs, use **[BioMate](https://www.biomate.ai?ref=kb&pkg=dada2)** — free to start.

▶ **[Open `dada2` on BioMate →](https://www.biomate.ai?ref=kb&pkg=dada2)**