Re: task pages: Moved many NGS packages from bio to bio-ngs
Hi Andreas,
On 24.09.17 14:06, Andreas Tille wrote:
> Hi Steffen,
>
> On Sun, Sep 24, 2017 at 11:24:49AM +0200, Steffen Möller wrote:
>> I had a bit of fun on that one:
>>
>> ...blend-med/tasks$ for i in bio bio-ngs; do egrep -v "^[ \t]*[X#]" $i |
>> sed -e 's/#.*$//' -e 's%//.*$%%' | grep -v '^\t'| egrep -v "^(
>> |Homepage|Lice|Pkg|Resp|Comm|Why|Lang|Rem|Meta|Task|Desc|WNPP)"| tr "
>> :,." "\n" | sort -u > $i.unique; done
>> ...blend-med/tasks$ join -v 2 bio.unique bio-ngs.unique
> Thanks for this code snippet. :-)
>
>> annovar <- non-free, was on wish list
>> hinge
>> scoary
>> umap
>>
>> It is far fewer than I had anticipated.
> Just add them to bio, please.
Done.
>
>>> BTW, I added the references to the UDD query that is used in the tasks
>>> pages generation code. So one further step is done to put them on the
>>> tasks pages. Unfortunately I was running (again) on a strange encoding
>>> problem with the UDD on my developer machine (luckily not on the
>>> production machine) which for the moment thwarts my attempt to finalise
>>> the web page. :-(
>> Big ":-)" on my side for every bit approaching some visibility of our
>> catalog references.
> Unfortunately there seems to be some data importer problem in UDD which
> causes a lack of descriptions for some packages. :-( Need to track this
> down first ...
>
Thanks.
As a bit of a sidenote - once we have more references to catalogs, we shall
use their references to workflows and also directly find references from CWL
descriptions of workflows to our packages to learn at what stage of a
bioinformatics workflow a package is used.
This information we can then use to
- give our task pages a bit of a graph structure in analogy to
domain-interaction graphs
- have a bit of a triage on maintaining the packages
This automated temporal ordering of packages (or of aggregations of these
packages as in the EDAM ontology) in software catalogs wrt the workflows
these appear in should be a nice student project, I tend to think. We
all need
to provide all these references, yet, though. Coming.
Best,
Steffen
Reply to: