Today I want to talk about installing Vicuna on Biolinux 8, if you were wondering why that should be something to write about it … well, it has been a hell of a task even for the people helping me. But first thing first what is Vicuna.
From their website (VICUNA) you can read that VICUNA is:
“VICUNA is a de novo assembly program targeting populations with high mutation rates. It creates a single linear representation of the mixed population on which intra-host variants can be mapped. For clinical samples rich in contamination (e.g., >95%), VICUNA can leverage existing genomes, if available, to assemble only target-alike reads. After initial assembly, it can also use existing genomes to perform guided merging of contigs. For each data set (e.g., Illumina paired read, 454), VICUNA outputs consensus sequence(s) and the corresponding multiple sequence alignment of constituent reads. VICUNA efficiently handles ultra-deep sequence data with tens of thousands fold coverage”.
Roughly 3 years ago we had VICUNA installed in our Biolinux system by our IT guy and one of the other PhD students, Gideon Mordecai, found this program very useful for his project and manages to assemble full viral genomes with this piece of software which are published here and here. So we were very happy until the IT left for a new job and long story short I was asked to help with the system that was up and running. This was a great opportunity for me to gain some extra skills and I thought ‘what can go wrong?’ …
I was so wrong…of course something had to go wrong.
After the Christmas holiday, two months after I took over the new system, we got back to the lab to find out that the hardware had failed and we had a technician coming in and installing a new version of Biolinux for us. At first I was happy because we would have the latest all singing, all dancing, Biolinux 8 installed in our system. Then it also scared the hell out of me, because they told me how expensive the system was and I didn’t want to screw it up. With Biolinux 8 finally installed I spent February and March updating programs and installing the programs needed, which wasn’t easy since I lacked confidence. But I managed. Since I know that lots of people had struggled to install VICUNA, I left that one for last. I contacted our old IT, and he was so helpful advising me on mounting the drives but about VICUNA he just couldn’t remember how he managed in the past.
Armed of a good dose of patience I tried to follow the instructions for external users. So first thing first you need NCBI-toolkit v 7.0.0 (but after experience you’ll just need the latest one!). Biolinux has a version of the NCBI tools included but, this is not the one you will need; you will need the ncbi_tools++. How do I know? I have bothered the Biolinux community which, by the way are absolutely amazing if you ever need some help they will be there guiding you and advising you. In this specific case I tried for a very long time to find the library of the tools installed and couldn’t find it … cause it wasn’t there. So I went on and installed a different tool, scared that I was going to kill the system (I shouldn’t have been afraid to mess it up since the system was new and should be working perfectly …. but I was being reminded daily of just how expensive the guy to install the system was).
I went on and install it in the share directory. Following the instructions from the manual, what I did was:
- Download NCBI toolkit from here: ftp://ftp.ncbi.nih.gov/toolbox/ncbi_tools++/CURRENT
- Untar the folder:
sudo tar –xvf ncbi_cxx--12_0_0.tar.gz
- Configure it
sudo ./configure –prefix=/home/share/Programs –with-optimization –with-mt—with-dll
This is where all the trouble started with all the possible errors. I’ve tried to install the ncbi_tools++ in the root and then as root and nothing, no way. Frustrated, I decided to ask online. I posted the question here: https://www.biostars.org/p/177810/#177897 and people where trying to help but none of the suggestions were working. And I was losing it (and hours/days of my PhD time). I really was *%$£. I posted for help on twitter and Tim from the UCT eResearch came to help. I know Tim from Cape Town since I am a user of the hex cluster and a registered PhD student of the University of Cape Town. I was very happy when he replied to me and even more since he went above and beyond to help me, following me every step and decoding every single error I was getting. He even installed Biolinux 8 in his machine to replicate it and manages to solve the problem for me.
So after a long rambling here is how to (actually) install VICUNA in Biolinux 8:
Biolinux8 doesn’t have a /opt directory so I created one:
cd / sudo mkdir opt/
Why did I do that? Well, Tim’s advice was that compiling application into /opt will give you more control over it. By the way, ncbi_tools lib takes approximately 50GB of space on your disk, so you want to make sure you have enough space available before you start.
Download and untar the ncbi_tools++
sudo tar –xvf ncbi_cxx--12_0_0.tar.gz
Configure ncbi_tools++. (This is what went wrong previously). The solution was to remove the –with-dll flag and this will make the libraries static instead of dynamic:
As a personal tip: create a specific folder for the libraries, then you’ll know exactly the path.
sudo mkdir ncbi_tools-lib
Then I configured it:
sudo ./configure –prefix=/opt/ncbi_tools-lib –with-optimization
sudo make sudo make install
NCBI_tools++ are now installed and configured.
Now for the beast which is VICUNA.
Downlad the tar from their website (VICUNA) and untar in the location you desire. I used the Programs folder I had created.
cd VICUNA bash #it is require to change environment
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/opt/ncbi_tools-lib cd src nano Makefile
Here you need to change the path of the library, for me it was /opt/ncbi_tools-lib. You also need to check the path for g++ and you can do so by typing “which g++” on the terminal and change it if necessary.
And VICUNA is up and running ^_^
last step, since ours is a shared system, I needed to change the permission of:
But that was it. It really took me a long time to complete this task and took lots of help from amazing people, as we say in Italy “Non e’ stata una passeggiata” (It wasn’t an easy walk) but we’ve made it.
What might make it easier in the future?
Currently if you try to contact the developer you will receive this error:
Delivery to the following recipient failed permanently:
Technical details of permanent failure:
We tried to deliver your message, but it was rejected by the server for the recipient domain returnmail.broadinstitute.org by mx2.broadinstitute.org. [126.96.36.199].
The error that the other server returned was:
550 5.1.1 <email@example.com>: Recipient address rejected: User unknown in local recipient table
And there was no response on research gate. If anyone involved with the original program can change the contact detail on the paper, this might help other users in the future with what is a very useful piece of software.
I really hope this blog post will help you install this software since I know it is very good for viral assembly. A massive thanks to the Biolinux support and Tim Carr from UCT eResearch.
Hope you enjoy and keep in touch,