Hi Jason and Hai,
I’m just writing to follow up on the sander API. The code works very well! One minor note was that I used chemistry.amber.NetCDFTraj.open_old() in ParmEd instead of chemistry.load_file. The object returned by load_file does not have the frame attribute.
Thanks so much. I think the sander API will be very interesting to work with.
- Lee-Ping
> On May 25, 2015, at 8:04 PM, Lee-Ping Wang <leeping.stanford.edu> wrote:
>
> Hai: Thanks a lot. I’ll see if I can get that to work once I get the data transferred, probably tomorrow because the data transfer is going to take a while. :)
>
> - Lee-Ping
>
>> On May 25, 2015, at 6:29 PM, Hai Nguyen <nhai.qn.gmail.com> wrote:
>>
>> .Lee-Ping: if you ever want to use `pytraj` with `sanderapi` and `ParmEd`,
>> you can in-the-fly (easily) play with single temperature frames from T-REMD
>> run without writing to temperature-trajectory first (thanks to cpptraj's
>> smart indexing).
>>
>> http://nbviewer.ipython.org/github/pytraj/pytraj/blob/master/note-books/load_remd_traj.ipynb
>>
>> Hai
>>
>> On Mon, May 25, 2015 at 4:43 PM, Lee-Ping Wang <leeping.stanford.edu> wrote:
>>
>>> Hi Jason,
>>>
>>> Thanks for the suggestion. I agree I could keep the water and use the
>>> sander API to compute the energies.
>>>
>>> - Lee-Ping
>>>
>>>> On May 25, 2015, at 12:12 PM, Jason Swails <jason.swails.gmail.com>
>>> wrote:
>>>>
>>>> On Mon, May 25, 2015 at 12:44 PM, Lee-Ping Wang <leeping.stanford.edu>
>>>> wrote:
>>>>
>>>>> Hi everyone,
>>>>>
>>>>> I’ve been running some replica exchange simulations in Amber. I’ve got
>>> a
>>>>> simple script for cpptraj (adapted from the tutorial at
>>>>> http://ambermd.org/tutorials/advanced/tutorial7/ <
>>>>> http://ambermd.org/tutorials/advanced/tutorial7/> ) for processing the
>>>>> output trajectories, resulting in “temperature-based” trajectories where
>>>>> the temperature is constant but the conformations jump around. The
>>>>> simulations are running very nicely; thanks for writing a great software
>>>>> package.
>>>>>
>>>>> My question is how I could get potential energies to “go with” these
>>>>> trajectories? I would like to use these potential energies as input to
>>> a
>>>>> WHAM / MBAR calculation to improve the estimation of thermodynamic
>>>>> properties at each temperature.
>>>>>
>>>>
>>>> The standard way is to feed those trajectories into sander, set imin=5
>>> to
>>>> process an input trajectory, and get energies that way. But I'll provide
>>>> an alternative I think you may find more amenabl
>>>> e:
>>>>
>>>> Use the sander API to compute single-point energies (and assuming you're
>>>> going to turn around and feed this directly into pymbar, you can do
>>>> everything you want in a single script). You can do something like this:
>>>>
>>>> import sander
>>>> import chemistry as chem
>>>>
>>>> parm = chem.amber.AmberParm('prmtop', 'inpcrd')
>>>> traj = chem.load_file('your_trajectory') # mdtraj or pytraj also works to
>>>> read trajectories
>>>>
>>>> input = sander.pme_input() # or use "gas_input(NUM)" where NUM is the igb
>>>> model for implicit solvent
>>>> input.cut = 9.0 # struct members have same name as mdin variables
>>>>
>>>> with sander.setup('prmtop', parm.coords, parm.box, input):
>>>> for i in range(traj.frame):
>>>> sander.set_positions(traj.coordinates(i))
>>>> ene, frc = sander.energy_forces()
>>>> # Go ahead and store your potential energy in a numpy array you
>>> set
>>>> up
>>>>
>>>> You can also use OpenMM instead of the sander Python API if you're more
>>>> comfortable with that.
>>>>
>>>>
>>>>
>>>>> My simple script is attached below.
>>>>>
>>>>> Thanks,
>>>>>
>>>>> - Lee-Ping
>>>>>
>>>>> parm ../prmtop
>>>>> ensemble ../remd.00/netcdf.00
>>>>> ensemble ../remd.01/netcdf.00
>>>>> ensemble ../remd.02/netcdf.00
>>>>> strip :WAT
>>>>>
>>>>
>>>> Obviously you can't calculate the full energies if you strip waters,
>>> but I
>>>> figured I'd point this out anyway.
>>>>
>>>> All the best,
>>>> 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
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sun May 31 2015 - 05:30:03 PDT
