Unicycler

Overview

Unicycler is a hybrid assembler specifically designed for bacterial genomes. It combines short Illumina reads with long reads from Oxford Nanopore or PacBio to produce complete, circularised chromosomes and plasmids. Unicycler first builds a short-read assembly graph using SPAdes, then uses long reads to resolve repeats and bridge gaps in the graph. It can also operate in short-read-only mode, where it functions as an optimised SPAdes pipeline with additional graph-cleaning heuristics. Unicycler assigns quality scores to completed replicons, making it straightforward to assess which contigs are fully resolved.

Installation

mamba install -c bioconda unicycler

Basic Usage

Perform a hybrid assembly using paired-end short reads and long reads.

# Hybrid assembly
unicycler -1 short_R1.fastq.gz -2 short_R2.fastq.gz \
  -l long_reads.fastq.gz \
  -o unicycler_output/ -t 8

Assemble a bacterial genome from short reads only.

# Short-read only
unicycler -1 short_R1.fastq.gz -2 short_R2.fastq.gz \
  -o unicycler_output/ -t 8

Key Parameters

Flag / option	Description
`-1` / `-2`	Input paired-end short-read FASTQ files (read 1 and read 2).
`-l`	Input long reads (Nanopore or PacBio) for hybrid assembly.
`-o`	Output directory for all assembly results.
`-t`	Number of CPU threads to use.
`--mode`	Assembly stringency: `conservative`, `normal` (default), or `bold`. Bold mode resolves more repeats but may introduce errors.
`--min_fasta_length`	Minimum contig length to include in the final output (default 100).
`--linear_seqs`	Expected number of linear sequences (default 0, assumes circular chromosomes and plasmids).
`--no_rotate`	Do not rotate completed circular sequences to start at a standard gene (e.g. dnaA).

Expected Output

Unicycler writes output to the specified directory:

assembly.fasta – the final assembly containing both complete and incomplete replicons in FASTA format.
assembly.gfa – the final assembly graph in GFA format, viewable in Bandage.
unicycler.log – a detailed log file recording every step of the assembly pipeline including SPAdes k-mer iterations, graph cleaning, and long-read bridging.

Each contig header in the FASTA output includes length, depth (coverage), and circularity status, allowing rapid identification of complete replicons.