-
Notifications
You must be signed in to change notification settings - Fork 12
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Remove Docker dependency #2
Comments
Hello |
I see, then perhaps I have some confusion. In the installation instructions for BiG-SCAPE under "Manual installation," there's no dependency on Docker, and I install the other dependencies via Conda, which works great in Google Colab. Since you said that CORASON also doesn't depend on Docker, is a similar "manual installation" possible for CORASON, and how could I do it? Currently when I clone the GutHub repo and run CCing @jorgecnavarrom in case he has some tips |
Its because BiG-SCAPE has two possible instalation forms, one is with conda and another one with docker, since CORASON is not python it has not the option to use conda. It will be hard to run a manual installation, for the dependencies, that is why conda and docker are valuable resources. For CORASON you need a linux or mac machine in order tat docker runs. Nevertheless I can run your data for you, If you whish. |
So in principle if those dependencies are installed (perhaps even through conda) it should be possible to run the CORASON perl scripts right? (e.g. https://github.com/nselem/corason/blob/master/CORASON/corason.pl) |
Yup that's what I'm thinking, and fortunately Conda does support Perl. I'll post some code for how to install those dependencies in a Perl virtual environment via Conda, but there will need to be a few changes to the master script that calls the others, right? |
Yes, we can try, the paths would be what we must change, but it would be great to have CORASON in CONDA too. |
This sounds great, and I don't think it should be too hard. After installing Conda, you can run the following commands in terminal:
This should get you started, but let me know if you run into any Conda problems. Do you also need to install any |
Ok, I will try, give me some days. :) Yes we need svg from cpan, basically, all needs are in the Dockerfile of the github repo. |
Ok, I just edited the above to include that. The code above should work, but if you run into any problems with cpan in conda, you can look here for troubleshooting You may need to put these two lines at the top of Perl scripts you want to run, but I'm not sure--it might just be enough to do
Please do let me know if I can help more with the Conda stuff. I know others use Conda + Perl a lot, so I'm sure we can get it working, and I think that would be great for users. |
Hi @nselem, |
Hello @rchurt,
I can try this as well, would love to have this tool in Conda. |
I followed these steps am at the point where I can actually run CORASON. Of note, the program run_corason points to docker, so, according to some documentation I found, instead we should be using corason.pl. Unfortunately, I am working with proprietary bacterial genomes and am not able to upload them to RAST for annotation. For my own research, I may have to build phylogenetic trees from the Big-Scape network data instead. |
Doesn't seem like I'll be able to test CORASON because the paremeters for "run_corason" and "corason.pl" are different, and I can't run corason.pl without RAST IDs. Is there a way to run it without RAST IDs? I'm trying to use the example dataset from the website. |
Hello jc, sorry, Im late with conda version, let me keep trying, you can use gbks. We can set a meeting next week and try do it togheter |
Hey @nselem, just wanted to see how this is going--did you manage to get it to work in Conda without Docker? |
Hello Curt |
Hello, here how you can use it with conda. (only tested with Rast files for now, gbk will be next step to be tried)
|
That's great news! I tried it out--below is my code and the output. It worked, but in the
Full output:
|
Can you show me the content of ctg2_515.query-output directory. Perhaps is the svg perl module (that I already have in my computer). Are the input files such as 501890_57.input empty or do they have some content?. If they have content, then we are only missing the drawing, that is the last step run by CORASON/3_Draw.pl. In fact your output shows or the modules |
Great, that worked. I had to manually install SVG and change the first line of Here's my code: https://colab.research.google.com/github/rchurt/rchurt.github.io/blob/master/CORASON.ipynb |
Glad it worked for you. Next step will be to making it work directly with gbk files. I dont know how to change SVG location without affecting docker users, it also worries me that the perl version that conda writes now 5.26.2 may change in future, and then users would need to change it any way. |
Hi, all, Alternatively, are there any instructions to make it run in Singularity? I have never used it before, but it seems to work to run docker images. Any help is greatly appreciated! |
Hello, Nelly |
@nselem Hi, nselem |
Hello,
I'd like to use Corason on a Google Colab virtual machine, but sadly they don't support Docker and have no plans to do so (googlecolab/colabtools#299). Unlike most packages that only use Docker for installation (e.g., BiG-SCAPE), I see that Corason actually uses Docker in its execution (e.g.,
docker run ...
in run_corason). Would it be very difficult to modify the scripts so that they can run without Docker?Thanks!
The text was updated successfully, but these errors were encountered: