hi
On Mon, May 22, 2017 at 10:39 AM, Brent Krueger <kruegerb.hope.edu> wrote:
> Thank you all who responded to my earlier note -- we have the sander api
> working for us in a basic sense.
>
> However, we are having trouble reproducing previous results. I suspect
> this is because we are comparing a sander_api energy using slightly
> different input parameters than our previously-computed sander energy,
> which is because the input data structure for the sander_api includes
input
> parameters that are a modest subset of the total number of sander input
> parameters.
>
> Two/three questions related to this:
>
> 1) Can I get the sander_api to dump something equivalent to the header in
> the mdout file that reports on the values for all of the input parameters?
Currently not, you need to update its code
>
> 2) Is it safe for me to assume that it is not too difficult to add a
> variable to the sander_api input data structure (e.g. dsum_tol)?
Yes, not that difficult.
>
> 3) And if so, what are all the pieces of code/subroutines that I need to
> touch to do this?
>
You can look at my change (adding ibelly option to sanderapi): Mostly copy
and paste but need to make sure
to declare variables in some files (otherwise it's easy to get segmentation
fault).
$ git log d33c1550a55a32c1ae785148e5655dbc3ea01a4c --name-only
AmberTools/src/pysander/sander/src/pysandermodule.c # --> only if you use
Python
AmberTools/src/pysander/sander/src/pysandermoduletypes.c # --> only if you
use Python
AmberTools/src/sander/interface.F90
AmberTools/src/sander/mdread1.F90
AmberTools/src/sander/mdread2.F90
AmberTools/src/sander/sander.h
Good luck
Hai
>
> Thanks,
> Brent
>
>
> On Tue, May 16, 2017 at 1:52 PM, Bill Ross <ross.cgl.ucsf.edu> wrote:
>
>> > could someone point me to the relevant lines of code that write amber
>> topology files?
>>
>> leap: unitio.c zUnitIOSaveAmberParmFormat
>>
>> Bill
>>
>>
>> On 5/16/17 9:15 AM, Brent Krueger wrote:
>> > Thank you Hai and Dan.
>> >
>> > My goal was to avoid ever writing the topology file to disk, but it
>> sounds
>> > like that is not possible. Probably, then, we'll just write the
topology
>> > file ourselves.
>> >
>> > I'm sure we can build this from the topology file format definition on
>> the
>> > amber site, but could someone point me to the relevant lines of code
that
>> > write amber topology files? Those will surely be a time-saver.
>> >
>> >
>> > Thanks,
>> > Brent
>> >
>> >
>> > On Tue, May 16, 2017 at 11:32 AM, Hai Nguyen <nhai.qn.gmail.com> wrote:
>> >
>> >> For now, I think the easiest way is to use sander Python API with
>> >> ParmEd's AmberParm. Under the hood, ParmEd save its object to prmtop
>> >> file and then passing to pysander.
>> >>
>> >> https://github.com/Amber-MD/pysander/blob/master/sander/__
>> init__.py#L275
>> >>
>> >> Of couse, we need to build ParmEd's AmberParm (either via tleap or
>> >> construct it in memory)
>> >>
>> >> Hai
>> >>
>> >> On Tue, May 16, 2017 at 11:03 AM, Daniel Roe <daniel.r.roe.gmail.com>
>> >> wrote:
>> >>> Hi,
>> >>>
>> >>> I don't think passing topology parameters directly is officially
>> >>> supported. A brief glance through sander.h shows that there is what
>> >>> looks like a topology structure and corresponding function calls:
>> >>>
>> >>> struct: prmtop_struct
>> >>> functions: sander_setup2 (2 versions)
>> >>>
>> >>> However, they don't appear to be documented or tested, so I would use
>> >>> these with extreme caution.
>> >>>
>> >>> FWIW, cpptraj makes use of the sander API by using the "officially"
>> >>> supported method of passing in a topology name. Since this only has
to
>> >>> be done once the actual overhead is minimal since the topology stays
>> >>> in memory after the setup call.
>> >>>
>> >>> -Dan
>> >>>
>> >>> On Tue, May 16, 2017 at 9:43 AM, Brent Krueger <kruegerb.hope.edu>
>> >> wrote:
>> >>>> We would like to use the sander API to do some energy evaluation of
>> some
>> >>>> simple systems. Part of the goal is to avoid lots of unnecessary
disk
>> >> I/O.
>> >>>> Therefore, we would like to create topology and coordinate
information
>> >>>> within our own program and keep it resident in memory. The
>> coordinates
>> >> are
>> >>>> easy enough, but I don't see anything in the manual about what data
>> >>>> structure holds the topology information and how we can access it.
>> All
>> >> of
>> >>>> the examples simply use the sander_setup_mm() function to read a
>> >> topology
>> >>>> file from disk.
>> >>>>
>> >>>> Maybe there are too many idiosyncrasies in the topology construction
>> >> and we
>> >>>> should just go ahead and use leap or parmed to do that. But our
>> >> situation
>> >>>> is very simple, We are modeling a crystal lattice, so there are no
>> bonds
>> >>>> just some ions with VDW parameters and charges. I feel like this is
>> >> simple
>> >>>> enough that we ought to be able to create this ourselves and pass it
>> to
>> >> the
>> >>>> sander API.
>> >>>>
>> >>>> Any information or examples would be appreciated. Oh, and we are
>> using
>> >> C,
>> >>>> but I suppose we could use anything else if that would make life
>> easier.
>> >>>>
>> >>>>
>> >>>> Thanks,
>> >>>> Brent
>> >>>>
>> >>>> --
>> >>>> _______________________________________________
>> >>>> Brent P. Krueger.....................phone: 616 395 7629
>> >>>> Professor.................................fax: 616 395 7118
>> >>>> Hope College..........................Schaap Hall 2120
>> >>>> Department of Chemistry
>> >>>> Holland, MI 49423
>> >>>> _______________________________________________
>> >>>> AMBER mailing list
>> >>>> AMBER.ambermd.org
>> >>>> http://lists.ambermd.org/mailman/listinfo/amber
>> >>>
>> >>>
>> >>> --
>> >>> -------------------------
>> >>> Daniel R. Roe
>> >>> Laboratory of Computational Biology
>> >>> National Institutes of Health, NHLBI
>> >>> 5635 Fishers Ln, Rm T900
>> >>> Rockville MD, 20852
>> >>> https://www.lobos.nih.gov/lcb
>> >>>
>> >>> _______________________________________________
>> >>> 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
>>
>
>
>
> --
> _______________________________________________
> Brent P. Krueger.....................phone: 616 395 7629
> Professor.................................fax: 616 395 7118
> Hope College..........................Schaap Hall 2120
> Department of Chemistry
> Holland, MI 49423
> _______________________________________________
> 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 Mon May 22 2017 - 19:30:03 PDT
