I agree 100% - Just look at the list of bugs we fixed from the release of
AMBER 12 to today. The GPU code was always labelled as experimental hence
the bugfixes etc. Anybody who doesn't update their AMBER installation
regularly is doing their users a massive disservice. To be honest I NEVER
trust centralized installations of AMBER and always compile and test my
own in my own home directory. You simply can't trust these centralized
computing facilities to keep things up to date. You could also take a look
at $AMBERHOME/logs and see if they ever bothered to runs the tests either
- I'd be amazed if they did.
So yes, I think papers should include the version number used to run the
simulations. Or at the very least the minimum. I try to do this with all
my papers, and include all the input files in the supplemental material -
that way the results can be exactly reproduced if ones wants to check.
Unfortunately I am now finding that a lot of journals despite letting you
upload tar or zip files containing the input as supplemental data never
actually publish it on their websites.
I'll put the input files for the lipid14 validation runs on my group's
website shortly and will send you the link.
All the best
Ross
On 1/31/14, 7:11 AM, "Jio M" <jiomm.yahoo.com> wrote:
>hi
>
>Its out of date indeed. I think all future publications should include
>not only AMBER version but the latest applied patches information as well
>;)
>this discussion gone somewhat out of topic I think, just a humble gentle
>reminder about Lipid 14 inputs; if possible.
>
>thanks
>
>
>
>
>
>On Friday, January 31, 2014 2:57 PM, Jason Swails
><jason.swails.gmail.com> wrote:
>
>On Fri, 2014-01-31 at 06:43 -0800, Jio M wrote:
>> Hi Jason
>>
>> Thanks, I can run this on my machine and I installed it recently:
>> $AMBERHOME/update_amber --version
>>
>> But this update_amber does not exist (in $AMBERHOME) in installed AMBER
>>version (some central facility). Is there some other way? I can see
>>though it was installed somewhat in June 2012! and I think update_amber
>>is new? Also not sure which bug fixes they have applied.
>
>If update_amber does not exist then it is very out-of-date.
>update_amber replaces patch_amber.py. You can go to $AMBERHOME and run
>the command "./patch_amber.py --patch-level" to get the same
>information, but the fact that update_amber is not present gives you
>your answer ;).
>
>
>HTH,
>Jason
>
>--
>Jason M. Swails
>BioMaPS,
>Rutgers University
>Postdoctoral Researcher
>
>
>_______________________________________________
>AMBER mailing list
>AMBER.ambermd.org
>http://lists.ambermd.org/mailman/listinfo/amber
>_______________________________________________
>AMBER mailing list
>AMBER.ambermd.org
>http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Jan 31 2014 - 09:30:02 PST
