Guix, CWL, ravanan and reproducibility talk at UCL

Guix, CWL, ravanan and reproducibility talk at UCL

[Arun Isaac]

Hi,

I am a postdoc at University College London (UCL). I have been working
on ravanan[1] for a few years now. ravanan is a Common Workflow
Language (CWL) runner that uses Guix to provide strong reproducibility
guarantees and content-addressed caching.

Yesterday, I gave a talk in my department at UCL on Guix, CWL, ravanan
and reproducibility. The slides are at
https://files.systemreboot.net/ucl-guix-ravanan-seminar.pdf

To the bioinformaticians reading this, it may be of interest that we
ported pggb (the pangenome building pipeline) from the original bash
script to CWL. Links to the original and the port below:

https://github.com/pangenome/pggb
https://github.com/arunisaac/pggb.cwl

Regards,
Arun

[1]: https://forge.systemreboot.net/ravanan/

Re: Guix, CWL, ravanan and reproducibility talk at UCL

[Konrad Hinsen]

Hi Arun,

> Yesterday, I gave a talk in my department at UCL on Guix, CWL, ravanan
> and reproducibility. The slides are at
> https://files.systemreboot.net/ucl-guix-ravanan-seminar.pdf

Thanks! "Pleasurable Reproducibility" is a term I'd like to see more
often!

Cheers,
  Konrad.
-- 
---------------------------------------------------------------------
Konrad Hinsen
Centre de Biophysique Moléculaire, CNRS Orléans
Synchrotron Soleil - Division Expériences
L'Orme des Merisiers
Départementale 128
91190 Saint Aubin, France
E-Mail: konrad DOT hinsen AT cnrs DOT fr
http://dirac.cnrs-orleans.fr/~hinsen/
ORCID: https://orcid.org/0000-0003-0330-9428
Mastodon: @khinsen@scholar.social
---------------------------------------------------------------------

Re: Guix, CWL, ravanan and reproducibility talk at UCL

[Ludovic Courtès]

Hi Arun,

Arun Isaac <arunisaac@systemreboot.net> skribis:

> I am a postdoc at University College London (UCL). I have been working
> on ravanan[1] for a few years now. ravanan is a Common Workflow
> Language (CWL) runner that uses Guix to provide strong reproducibility
> guarantees and content-addressed caching.
>
> Yesterday, I gave a talk in my department at UCL on Guix, CWL, ravanan
> and reproducibility. The slides are at
> https://files.systemreboot.net/ucl-guix-ravanan-seminar.pdf

That’s a very nice slide deck, thanks for sharing!

I would encourage you to turn your talk in a post at
<https://hpc.guix.info/blog> and/or in a refereed paper.

> To the bioinformaticians reading this, it may be of interest that we
> ported pggb (the pangenome building pipeline) from the original bash
> script to CWL. Links to the original and the port below:

Do you have experience running existing CWL bioinfo pipelines “from out
there” with ravanan?

Thanks,
Ludo’.

Re: Guix, CWL, ravanan and reproducibility talk at UCL

[Arun Isaac]

Hi Ludo,

> I would encourage you to turn your talk in a post at
> <https://hpc.guix.info/blog> and/or in a refereed paper.

Yes, certainly on the cards! Sorry I've been promising this for many
months now but haven't delivered yet. But, will do.

>> To the bioinformaticians reading this, it may be of interest that we
>> ported pggb (the pangenome building pipeline) from the original bash
>> script to CWL. Links to the original and the port below:
>
> Do you have experience running existing CWL bioinfo pipelines “from out
> there” with ravanan?

No, I haven't tried yet. But, that's a good idea. I will try it out.

Regards,
Arun

Re: Guix, CWL, ravanan and reproducibility talk at UCL

[Reza Housseini]

Hi Arun

Thanks for sharing! Do you have any experience running this from Python?
E.g. generating the CWL files in Python? And do I understand corecctly
that ravanan is not tied to CWL and/or guix?

Best,
Reza

> Hi,
>
> I am a postdoc at University College London (UCL). I have been working
> on ravanan[1] for a few years now. ravanan is a Common Workflow
> Language (CWL) runner that uses Guix to provide strong reproducibility
> guarantees and content-addressed caching.
>
> Yesterday, I gave a talk in my department at UCL on Guix, CWL, ravanan
> and reproducibility. The slides are at
> https://files.systemreboot.net/ucl-guix-ravanan-seminar.pdf
>
> To the bioinformaticians reading this, it may be of interest that we
> ported pggb (the pangenome building pipeline) from the original bash
> script to CWL. Links to the original and the port below:
>
> https://github.com/pangenome/pggb
> https://github.com/arunisaac/pggb.cwl
>
> Regards,
> Arun
>
> [1]: https://forge.systemreboot.net/ravanan/

Re: Guix, CWL, ravanan and reproducibility talk at UCL

[Arun Isaac]

Hi Reza,

> Do you have any experience running this from Python?
> E.g. generating the CWL files in Python?

I normally generate my CWL files using ccwl:
https://ccwl.systemreboot.net/

Several other CWL generators exist as well:
https://www.commonwl.org/tools/#converters

And you can also write CWL by hand.

> And do I understand corecctly that ravanan is not tied to CWL and/or
> guix?

I'm not sure I understand what you mean. ravanan is a CWL runner that
uses Guix. So, it is tied to both CWL and Guix.

Regards,
Arun