Let's start with the publications Re: Fwd: [bio-tools/biotoolsRegistry] bio.tools COVID-19 tools list (#505)
Virologists are rarely very creative when in a hurry. They want to know
something and have read in another paper how this worked for another
group. So they want that, too. And they want that quickly.
I propose that we take some of these papers and demonstrate that we can
repeat the presented analysis. Some differences are expected since we
will have newer data, but hey, that is where it all starts to become
interesting. We could even come up with a perpetual version of that
paper, i.e. a website that is auto-updated with figures from the paper,
just with newer data. Actually, I think I like that. Everyone pick a
paper to their liking and off we go, quickly discussing overlaps in the
selected tools.
We never solved the problem how to best handle scientific data in
Debian. This is something we should think about.
To myself I'll assign a work from a group geographically close to me
that has repeated its success with the first SARS virus. They have now
determined the 3D structure of the SARS-CoV-2 protease, i.e. the enzyme
that cuts what is read from its RNA genome into functional pieces.
_L. Zhang_, _D. Lin_, _X. Sun_, U. Curth, C. Drosten, L. Sauerhering, S.
Becker, K. Rox & _R. Hilgenfeld:_
Crystal structure of SARS-CoV-2 main protease provides a basis for
design of improved *α*-ketoamide inhibitors.
/Science/ (2020). doi: 10.1126/science.abb3405. [Epub ahead of print]
http://www.biochem.uni-luebeck.de/public/publications/Zhang_et_al_Science_2020.pdf
I suggest we start with a package providing that data, and another one
with sequence data, and then we start playing;
a) show how to 3D-inspect the data on the local computer
b) set up a quick docking experiment with AutoDock to redock the
compound they have
c) get a sequence alignment for the protease and look at the most
conserved residues (those are the important ones)
d) look at mutual information between sequence variants
e) show how to play with b-e, say with biopython or BALL - we just
kicked the AutoDockToolkit out, which is unfortunate, Rosetta is very
non-free, Pymol maybe.
Should someone else want to do this - I happily train a volunteer.
My next stop may then be viral genome assembly from NGS data, -> Literature.
Then maybe the representation of the viral genome. Would now start with
a literature search.
Would you agree that the literature-driven approach is worthwhile? This
should also bring is in close contact with the respective authors from
whom we may have to collect the data if not available online (which
should not be the case these days).
Steffen
On 25.03.20 17:51, Michael Crusoe wrote:
FYI
---------- Forwarded message ---------
From: *Hans Ienasescu* <notifications@github.com
<mailto:notifications@github.com>>
Date: Wed, Mar 25, 2020 at 5:50 PM
Subject: [bio-tools/biotoolsRegistry] bio.tools COVID-19 tools list (#505)
This issue is a placeholder for a discussion related to the
bio.tools COVID-19 related tools list.
There is a *COVID-19* related subdomain in bio.tools at:
*https://covid-19.bio.tools <https://covid-19.bio.tools>*
*Everyone* is encouraged to populate the tools list at the bio.tools
subdomain above.
In order to populate the subdomain you can:
* *post a link to a bio.tools entry in this thread as a comment*
* *Tag tools in bio.tools with the |COVID-19| collection*
* Email registry-support@elixirmail.cbs.dtu.dk
<mailto:registry-support@elixirmail.cbs.dtu.dk> with tool
recommendations (*less recommended*, /only if you don't have a
bio.tools or GitHub account/)
* Post a link to a tool not in bio.tools (*less recommended*, /might
take a long time until the tool actually gets into bio.tools/)
If you post a tool here or tag tools with the |COVID-19| collection, I
(and other willing folks) will update the *https://covid-19.bio.tools
<https://covid-19.bio.tools>* subdomain as soon as possible.
*Please only post or add tools that are of real relevance to
COVID-19!* What is of /real relevance/ is open to discussion and we
should have that discussion, but let's not spam this thread with
arguments on what is relevant.
--
Michael R. Crusoe
Reply to: