Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
    Johannes Dröge
    @fungs
    this is some module which can create a python argparse (option parsing) object directly from a YAML or JSON file: https://github.com/fmenabe/python-clg
    it might be useful
    Michael Barton
    @michaelbarton
    Sounds like Peter’s talk at BOSC went well - https://twitter.com/GigaScience/status/619823935933284352
    Andreas Bremges
    @abremges
    Yes, from what I've heard from Scott Edmunds & Peter himself
    Definitely the right place to present bioboxes
    Michael Barton
    @michaelbarton
    I have got a prototype branch working on the biobox cli.
    This allows the following command biobox short_read_assembler bioboxes/megahit --verify
    This verifies that a short_read_assembler biobox conforms to the spec.
    I’m most excited about this because building and testing new biobox docker containers is not easy, but this makes it much more so.
    Albert Vilella
    @avilella
    I may be late in finding out about this DinD approach but here it goes: https://twitter.com/ProfParmer/status/651835246158024704
    Paolo Di Tommaso
    @pditommaso
    It looks this approach has a serious security problem
    Michael Barton
    @michaelbarton
    I agree with the blog post, providing access to the docker socket is still a security concern. It would be nice if there was a better way to support Docker in Docker or at least some variant.
    NERSC have created ‘shifter’ which seems to provide a safer way to do this in a HPC environment - https://www.nersc.gov/news-publications/nersc-news/nersc-center-news/2015/shifter-makes-container-based-hpc-a-breeze/
    Paolo Di Tommaso
    @pditommaso
    It looks an interesting project, and I suppose that this kind of "hybrid" approach (i.e. Docker + custom tools) is the most likely to take place in the HPC
    Daniel Schober
    @DSchober
    Hi there, at the moment I see very few bioboxes available and their presentation as a simple list might be sufficient for the moment, but if this repository grows, you will need a way to search for particulat boxes. Is there anything planned yet to group/classify and annotate biobox functionalities in a standardized way?
    Johannes Dröge
    @fungs
    @DSchober you are right, currently we are only maintaining a plain list. Bioboxes tries to follow the approach "by users for users" and thus we will have to come up with a more accessible and detailed index data structure which users can contribute to. Since our central hub currently is the GitHub project, I think that YAML/JSON files (over HTTPS with clients) will be a good start.
    Michael Barton
    @michaelbarton
    @fungs There is a YAML document created by @pbelmann here -https://github.com/bioboxes/data/blob/master/images.yml
    @DSchober We have discussed using signatures as a way to classify bioboxes - bioboxes/rfc#9. This would be a way to classify each biobox based on the inputs it accepts and the outputs it returns.
    Johannes Dröge
    @fungs
    Hi Michael, the index look but I think it lacks the bioboxes.org interface which each tasks implements
    Andreas Bremges
    @abremges
    Idea – we could add this gitter room to the bioboxes.org homepage using the new sidecar widget: https://sidecar.gitter.im/ & https://github.com/gitterHQ/sidecar
    Michael Barton
    @michaelbarton
    @abremges if you’d be willing to add this to the site, I think this would be useful.
    Michael Barton
    @michaelbarton
    For those involved in CAMI, did you benchmark the LSA/eigengenomes paper? Does this fit into the binning part of the challenge?
    Andreas Bremges
    @abremges
    We did not receive an entry from the authors, and the paper appeared after the challenge closed (if I remember correctly). From my point of view, it's an exciting combination of binning/assembly, but at least the initial binning step could be benchmarked on its own.
    And I'll take a look at the widget, but I'm not too familiar with such stuff...
    Michael Barton
    @michaelbarton
    I have found a bug where the CLI does not work on OSX - bioboxes/rfc#186
    Jonathan Narvaez
    @jnarvaezp
    Hi, Anything here?
    Michael Barton
    @michaelbarton
    Hi Jonathan, what can I do for you?
    Jonathan Narvaez
    @jnarvaezp
    hi, i try to replicate a pipeline, i port multiple tools to biobox, but i don't understand, how to make the structure of a DAG
    Michael Barton
    @michaelbarton
    The DAG?
    Creating the biobox?
    Jonathan Narvaez
    @jnarvaezp
    yes, i have a pipeline Directed acyclic graph (DAG), how implement this pipeline with biobox.
    Michael Barton
    @michaelbarton
    I think you’ll have have to ensure you can build your pipeline inside a Docker container first. Have you gotten to this point already?
    Jonathan Narvaez
    @jnarvaezp
    i have planned to use bioboxes to run a pipeline in different nodes, i was develop a strategy using distributed systems for processing pipelines, but i try to integrate bioboxes in my tool.
    Michael Barton
    @michaelbarton
    I see. I think I understand.
    At the moment we have biobox for a couple of applications: assembly, binning and profiling.
    To a run a biobox you will need to create a biobox.yaml for each type and pass this to biobox Docker image when you start it.
    In general, if you can run Docker on your nodes, you should be able to run bioboxes as they are a specific type of Docker image
    Jonathan Narvaez
    @jnarvaezp
    Ok, thanks for your help, i try to port hmmer, blast, cufflinks. when images are ready, i do commit.
    Philip Blood
    @pdblood
    @pbelmann @michaelbarton Hi, I went to pull some bioboxes from the website, and the links take me to my home page on dockerhub
    Philip Blood
    @pdblood
    Also, I've seen some posts by the Galaxy community about BioContainers: https://github.com/BioContainers
    Are you familiar with this effort, and if so what are your thoughts?
    pbelmann
    @pbelmann

    @pdblood thanks for telling us that, we are currently working on a website update. see here: bioboxes/rfc#203 . We will fix the links as soon as possible.

    Regarding BioContainers I would say that it is a great effort for integrating bioinformatic tools in docker images. In my opinion it has not the same goal as bioboxes. Bioboxes aim is to make bioinformatic software interchangeable and by doing that making it easier using multiple containers or images of one type (assemblers/profilers and so on).

    Michael Barton
    @michaelbarton
    @pdblood Thanks for all your help this weekend Phil. With regard to your comment above, I believe these links are all fixed now.
    KoGro
    @KoGro
    Hi there! Is there any overview/documentation of all the possible options for the biobox.yaml files? E.g. is it possible to list more than one input file. Or how would I tell velvet (for example) that I have paired end reads? Thanks!
    pbelmann
    @pbelmann
    @KoGro See here for all possible values for a biobox.yaml https://github.com/bioboxes/rfc/blob/master/container/short-read-assembler/rfc.mkd
    KoGro
    @KoGro
    Thanks, @pbelmann !
    Francesco Beghini
    @fbeghini
    Hi everyone! Is this channel still active?
    anonymous
    @naveenmeena584
    hello