Aw: Re: packages for analyses of DNA structural variations
> Gesendet: Montag, 12. Juli 2021 um 13:18 Uhr
> Von: "Nilesh Patra" <nilesh@debian.org>
> An: debian-med@lists.debian.org
> Betreff: Re: packages for analyses of DNA structural variations
>
> On 11/07/21 10:04 PM, Steffen Möller wrote:
> > Hello,
> >
> > I went a bit over packages listed in the "genome structure" tab
> > https://docs.google.com/spreadsheets/d/1tApLhVqxRZ2VOuMH_aPUgFENQJfbLlB_PFH_Ah_q7hM/edit#gid=701092179
> >
> > and have
>
> > surpyvor
>
> Did the following things:
>
> * Add missing build-depends and depends -- please build in a clean chroot :-)
> * Updated d/u/metadata
>
> However, this depends on the upsetplot package[1], which is not in the
> archive -- can you please package this?
Sorry, yes. Repeated offender, I admit.
>
> I can't even run a --help command without this package installed.
>
> $ grep -irn upsetplot
> surpyvor/plots.py:6:from upsetplot import plot as upsetplot
> surpyvor/plots.py:50:def upset_plot(upsets, outname="UpSetPlot.png"):
> surpyvor/plots.py:51: upsetplot(upsets, sort_by='cardinality')
> surpyvor/parse_arguments.py:172: default="UpSetPlot.png")
>
> > tandem-genotypes considered worthy for sponsoring.
> >
>
> * Updated to new upstream
> * Upstream seems to vendor tests, but it does not seem to vendor those
> in the tarball. So for now, I pulled those manually and did a multi-orig
> tarball solution for the same.
> Maybe we can convince upstream to vendor these in the next release
I am in contact with upstream about this.
> * Added autopkgtests
> * minor fixes in d/control
> * Fixed shebang
>
> I also went ahead and added myself to uploaders, hoping that to be OK
> Since I was happy with the package, I uploaded.
This is nice to hear. Thank you for making this package shine.
>
> Please consider enabling Salsa-CI for any packages you work on, I did it
> for thesew two for now
> and as I said earlier, I also added this in the inject-into-salsa-git
> script[2] so this is done by default please consider to pull my changes
> from there
Ah, this explains it. I had used that script and had considered that settled because of that, but likely my local version was outdated. I'll fix that.
> With this, I have 32 packages in new, and I really do hope that FTP
> masters get to it soonish.
>
> Thanks a lot for having worked on these packages, and please consider to
> do the needful to get surpyvor in as well
>
> [1]: https://github.com/jnothman/UpSetPlot
> [2]: https://salsa.debian.org/med-team/community/helper-scripts/-/commit/ae237adad67d6501459706f53eba89992b21e694
We almost got the nano-snakemake pipeline complete. A bit of a bummer is that pbsv is non-free, despite the declared BSD license in the repository. Upstream points to conda as their sole distribution channel - you can read this up in the issue I referenced in my previous email. Need to think a bit more what to do about this. One solution could be to prepare a binary-package that we derive from conda. I do not think that I want that, though. Or we get conda into Debian and execute that as a post-install script of a pbsv-via-conda package.
Best,
Steffen
Reply to: