Skip to content


Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time

Table of Contents


Redundans pipeline assists an assembly of heterozygous genomes.
Program takes as input assembled contigs, sequencing libraries and/or reference sequence and returns scaffolded homozygous genome assembly. Final assembly should be less fragmented and with total size smaller than the input contigs. In addition, Redundans will automatically close the gaps resulting from genome assembly or scaffolding.

The pipeline consists of several steps (modules):

  1. de novo contig assembly (optional if no contigs are given)
  2. redundancy reduction: detection and selective removal of redundant contigs from an initial de novo assembly
  3. scaffolding: joining of genome fragments using paired-end reads, mate-pairs, long reads and/or reference chromosomes
  4. gap closing: filling the gaps after scaffolding using paired-end and/or mate-pair reads

Redundans is:

  • fast & lightweight, multi-core support and memory-optimised, so it can be run even on the laptop for small-to-medium size genomes
  • flexible toward many sequencing technologies (Illumina, 454, Sanger, PacBio & Nanopore) and library types (paired-end, mate pairs, fosmids, long reads)
  • modular: every step can be omitted or replaced by other tools
  • reliable: it has been already used to improve genome assemblies varying in size (several Mb to several Gb) and complexity (fungal, animal & plants)

For more information have a look at the documentation, poster, publication, test dataset or manual.


Redundans uses several programs (all provided within this repository):

On most Linux distros, the installation should be as easy as:

git clone --recursive
cd redundans && bin/

If it fails, make sure you have below dependencies installed:

  • Python 2.7 or 2.6
  • Perl [SSPACE3]
  • make, gcc & g++ [BWA & LAST] ie. sudo apt-get install make gcc g++
  • zlib including zlib.h headers [BWA] ie. sudo apt-get install zlib1g-dev
  • optionally for plotting numpy and matplotlib ie. sudo -H pip install -U matplotlib numpy

For user convenience, we provide UNIX installer and Docker image, that can be used instead of manually installation.

Unofficial conda package

If you are familiar with conda, this will be by far the easiest way of installing redundans:

# create new Python2 environment
conda create -n redundans python=2.7
# activate it
conda activate redundans
# and install redundans
conda install -c genomedk redundans 

Note, this is unofficial channel and may not be completely up-to-date with this repo.

UNIX installer

UNIX installer will automatically fetch, compile and configure Redundans together with all dependencies. It should work on all modern Linux systems, given Python 2.7, commonly used programmes (ie. wget, curl, git, perl, gcc, g++, ldconfig) and libraries (zlib including zlib.h) are installed.

source <(curl -Ls

Docker image

First, you need to install docker: wget -qO- | sh
Then, you can run the test example by executing:

# process the data inside the image - all data will be lost at the end
docker run -it -w /root/src/redundans lpryszcz/redundans ./ -v -i test/{600,5000}_{1,2}.fq.gz -f test/contigs.fa -o test/run1

# if you wish to process local files, you need to mount the volume with -v
## make sure you are in redundans repo directory (containing test/ directory)
docker run -v `pwd`/test:/test:rw -it lpryszcz/redundans /root/src/redundans/ -v -i test/*.fq.gz -f test/contigs.fa -o test/run1

Docker images are very handy, but they have certain limitation. The most annoying for me is the lack of autocompletion, unless you specify the path in host and container in the exactly same manner as in the example above. In addition, the volume needs to be mounted every time, leading to a bit complex commands.

Running the pipeline

Redundans input consists of any combination of:

  • assembled contigs (FastA)
  • paired-end and/or mate pairs reads (FastQ*)
  • long reads (FastQ/FastA*) - both PacBio and Nanopore are supported
  • and/or reference chromosomes/contigs (FastA).
  • gzipped files are also accepted.

Redundans will return homozygous genome assembly in scaffolds.filled.fa (FastA).
In addition, the program reports statistics for every pipeline step, including number of contigs that were removed, GC content, N50, N90 and size of gap regions.


For the user convenience, Redundans is equipped with a wrapper that automatically estimates run parameters and executes all steps/modules. You should specify some sequencing libraries (FastA/FastQ) or reference sequence (FastA) in order to perform scaffolding. If you don't specify -f contigs (FastA), Redundans will assemble contigs de novo, but you'll have to provide paired-end and/or mate pairs reads (FastQ). Most of the pipeline parameters can be adjusted manually (default values are given in square brackets []):
HINT: If you run fails, you may try to resume it, by adding --resume parameter.

  • General options:
  -h, --help            show this help message and exit
  -v, --verbose         verbose
  --version             show program's version number and exit
  -i FASTQ, --fastq FASTQ
                        FASTQ PE / MP files
  -f FASTA, --fasta FASTA
                        FASTA file with contigs / scaffolds
  -o OUTDIR, --outdir OUTDIR
                        output directory [redundans]
  -t THREADS, --threads THREADS
                        no. of threads to run [4]
  --resume              resume previous run
  --log LOG             output log to [stderr]
  • Reduction options:
  --identity IDENTITY   min. identity [0.51]
  --overlap OVERLAP     min. overlap  [0.80]
  --minLength MINLENGTH
                        min. contig length [200]
  --noreduction         Skip reduction
  • Short-read scaffolding options:
  -j JOINS, --joins JOINS
                        min pairs to join contigs [5]
  -a LINKRATIO, --linkratio LINKRATIO
                        max link ratio between two best contig pairs [0.7]
  --limit LIMIT         align subset of reads [0.2]
  -q MAPQ, --mapq MAPQ  min mapping quality [10]
  --iters ITERS         iterations per library [2]
  --noscaffolding       Skip short-read scaffolding
  • Long-read scaffolding options:
  -l LONGREADS, --longreads LONGREADS
                        FastQ/FastA files with long reads
  --identity IDENTITY   min. identity [0.51]
  --overlap OVERLAP     min. overlap  [0.80]
  • Reference-based scaffolding options:
  -r REFERENCE, --reference REFERENCE
                        reference FastA file
  --norearrangements    high identity mode (rearrangements not allowed)
  --identity IDENTITY   min. identity [0.51]
  --overlap OVERLAP     min. overlap  [0.80]
  • Gap closing options:
  --iters ITERS         iterations per library [2]

Redundans is extremely flexible. All steps of the pipeline can be ommited using: --noreduction, --noscaffolding and/or --nogapclosing parameters.

Test run

To run the test example, execute:

./ -v -i test/*_?.fq.gz -f test/contigs.fa -o test/run1

# if your run failed for any reason, you can try to resume it
rm test/run1/_sspace.2.1.filled.fa
./ -v -i test/*_?.fq.gz -f test/contigs.fa -o test/run1 --resume

# if you have no contigs assembled, just run without `-f`
./ -v -i test/*_?.fq.gz -o test/run.denovo

Note, the order of libraries (-i/--input) is not important, as long as read1 and read2 from each library are given one after another i.e. -i 600_1.fq.gz 600_2.fq.gz 5000_1.fq.gz 5000_2.fq.gz would be interpreted the same as -i 5000_1.fq.gz 5000_2.fq.gz 600_1.fq.gz 600_2.fq.gz.

You can play with any combination of inputs ie. paired-end, mate pairs, long reads and / or reference-based scaffolding, for example:

# reduction, scaffolding with paired-end, mate pairs and long reads, and gap closing with paired-end and mate pairs
./ -v -i test/*_?.fq.gz -l test/pacbio.fq.gz test/nanopore.fa.gz -f test/contigs.fa -o test/run_short_long

# scaffolding and gap closing with paired-end and mate pairs (no reduction)
./ -v -i test/*_?.fq.gz -f test/contigs.fa -o test/run_short-scaffolding-closing --noreduction

# reduction, reference-based scaffolding and gap closing with paired-end reads (--noscaffolding disables only short-read scaffolding)
./ -v -i test/600_?.fq.gz -r test/ref.fa -f test/contigs.fa -o test/run_ref_pe-closing --noscaffolding

For more details have a look in test directory.


If you have any issues or doubts check documentation and FAQ (Frequently Asked Questions). You may want also to sign to our forum.


Leszek P. Pryszcz and Toni Gabaldón (2016) Redundans: an assembly pipeline for highly heterozygous genomes. NAR. doi: 10.1093/nar/gkw294