Overview
velvet is a software package that can manipulate de Bruijn graphs for genomic sequence assembly. A de Bruijn graph is a compact representation based on short words (k-mers) that is ideal for high coverage, very short read (25–50 bp) data sets. velvet works with short reads and produce contigs of significant length.
velvet is available on github.
To learn more about velvet you can read the associated paper.
Downloading and installing velvet
You can download and compile velvet using the following commands:
git clone https://github.com/dzerbino/velvet
cd velvet
make
If you want to install velvet with OpemMP support to enable multi-threading you can compile velvet with:
make 'OPENMP=1'
To verify that your installation supports OpenMP you can run ./velveth and see under compilation settings:
Version 1.2.10
Copyright 2007, 2008 Daniel Zerbino (zerbino@ebi.ac.uk)\This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.Compilation settings:
CATEGORIES = 2
MAXKMERLENGTH = 31
OPENMP
…
Running the included velvet tests
The velvet package includes a number of tests to verify that the software runs correctly. To run the tests follow these commands”
cd tests
./run-tests.sh
Dockerizing velvet
velvet can be configured to run from within a docker container. There are a number of ways to do this. The following dockerfile shows how to create a docker image that loads velvet to a docker container from a local copy, compile it within the container, and run the velvet tests.
FROM gcc:5.4
RUN mkdir -p /app/velvet
ADD Makefile /app/velvet
RUN mkdir /app/velvet/src
ADD src /app/velvet/src
RUN mkdir /app/velvet/tests
ADD tests /app/velvet/tests
RUN mkdir /app/velvet/third-party
ADD third-party /app/velvet/third-party
WORKDIR /app/velvet
RUN make
WORKDIR /app/velvet/tests
ENTRYPOINT ["./run-tests.sh"]
Make sure you save the above dockerfile in the velvet root directory and then run
docker build velvet:latest .
docker run velvet:latest
velvet drawbacks
The velvet software package is fairly old and not very well maintained. At the time of writing, the latest commit in https://github.com/dzerbino/velvet was on Aug 15th, 2014. A major drawback of de novo genome assemblers, including but not specific to velvet, is that they require a lot of memory.
Resources
- https://github.com/dzerbino/velvet
- https://www.ebi.ac.uk/~zerbino/velvet/
- https://holtlab.net/tag/velvet/