Re: [AMBER] Calculation of the number of internal waters during md simulation

From: Daniel Roe <daniel.r.roe.gmail.com>
Date: Fri, 6 Feb 2015 10:52:11 -0700

> I think multhist has not been included yet to the amber-12 which I'm using

I highly recommend upgrading to AmberTools 14, which is free. It
includes many updates and enhancements over Amber 12, which is several
years old at this point.

On Fri, Feb 6, 2015 at 2:05 AM, James Starlight <jmsstarlight.gmail.com> wrote:
>
> [reference ./protein.pdb]
> Error: ./protein.pdb: No frames read. atom=4542 expected 46488.

Cpptraj requires topology information to load coordinates, or to put
it another way there must be a 'parm' that can be used with any given
'trajin'. So what you really want is something like:

parm ../protein.pdb [REFPARM]
reference ../protein.pdb parm [REFPARM]

Even though 'protein.pdb' contains both topology and coordinate
information, cpptraj doesn't assume you want to use them together, the
reason being that you may want to e.g. use that PDB file with a
matching Amber topology file (to get things like bonds, charges, etc).

> (2)
> # here I chose grid dimensions from the XYX of the box vectors proding mask
> of the solvent
> grid grid.dat 60 1 80 1 100 1 :WAT.O*
>
> 1- should I add something else to this conditions in case of my task?
> 2- How it would be better to compare several grid.dat output files produced
> for several trajectories?

You may want to just focus your grid around the molecule of interest
(in your case I think it was :MOL). The simplest way to do that in my
opinion is to upgrade to cpptraj from AmberTools 14 and use the
'bounds' command, which will provide minimum and maximum XYZ values
for a selection of atoms that you can then use to create a grid. Also,
for something like this its probably best to make sure your molecule
of interest and the grid itself are centered at the origin. See the
manual for relevant keywords.

Hope this helps,

-Dan

>
>
> James
>
> 2015-02-06 9:23 GMT+01:00 James Starlight <jmsstarlight.gmail.com>:
>
>> Thanks Dan,
>>
>> I think multhist has not been included yet to the amber-12 which I'm using
>>
>> ...
>> INPUT: Reading Input from file water.in
>> [parm ./top_all/7D4-androsta.top]
>> [trajin ./tr_all/7D4-androsta.mdcrd]
>> [7D4-androsta.mdcrd] contains 5191 frames.
>> [watershell :104 W1_apo_DRY.dat :WAT.O* lower 3 upper 5]
>> [multihist W1[*] out W1.hist.agr bins 50]
>> Warning: Unknown Command multihist.
>>
>>
>> James
>>
>> 2015-02-05 17:06 GMT+01:00 Daniel Roe <daniel.r.roe.gmail.com>:
>>
>>> The 'grid' command will allow you to calculate water occupancy (not
>>> yet normalized, although this option is in the next cpptraj release)
>>> which you can compare between different simulations. However, it's
>>> probably best to rms-fit the solute (i.e. the protein, not the ligand)
>>> to a common reference prior to 'grid' to remove rotation/translation
>>> of the solute. In other words, you want to view water occupancy with
>>> your protein as the frame of reference.
>>>
>>> -Dan
>>>
>>> On Thu, Feb 5, 2015 at 2:24 AM, James Starlight <jmsstarlight.gmail.com>
>>> wrote:
>>> > (+) How do you think will it be better to use grid command (in
>>> comparison
>>> > to watershell) from the cpptraj to compare probability of the water
>>> > distribution throughout the protein interior during several md runs? In
>>> > general watershell gave me satisfied results but I'd like to obtain more
>>> > statistical-relevant distribution avoiding with the choosing of proper
>>> > cut-offs.
>>> >
>>> > James
>>> >
>>> > 2015-02-04 10:30 GMT+01:00 James Starlight <jmsstarlight.gmail.com>:
>>> >
>>> >> Could you also provide me with the link where is possible to download
>>> >> multhist tool? As I realize it might be some python script which I had
>>> not
>>> >> find in my amber12. IS it possible to plot such multiple graph values
>>> >> agains time prolongation using xmgrace?
>>> >>
>>> >> Thanks,
>>> >>
>>> >> James
>>> >>
>>> >> 2015-02-03 17:38 GMT+01:00 James Starlight <jmsstarlight.gmail.com>:
>>> >>
>>> >>> Thanks Dan!
>>> >>>
>>> >>> I'm not sure only if I selected mask for the solvent properly but it
>>> >>> produced good graph. Taking into account that :MOL is the ligand
>>> within the
>>> >>> cavity and I'd like to estimate water influx exactly into the cavity
>>> during
>>> >>> md simulation what reasonable lower and upper cutoffs might be?
>>> >>>
>>> >>> watershell :MOL W1.dat :WAT lower 3.0 upper 5.0
>>> >>>
>>> >>> PARM [protein.parm7]: Setting up 1 actions.
>>> >>> 0: [watershell :MOL W1.dat :WAT lower 3.0 upper 5.0]
>>> >>> Mask [:MOL] represents 46 atoms
>>> >>> Mask [:WAT] represents 25560 atoms
>>> >>> WATER SHELL: Output to W1.dat
>>> >>> The first shell will contain water < 3.000 angstroms from
>>> >>> the solute; the second shell < 5.000 angstroms...
>>> >>> The solute atoms are :290
>>> >>> The solvent atoms are :396-8914,:8915.O,:8915.H1,:8915.H2
>>> >>>
>>> >>> James
>>> >>>
>>> >>>
>>> >>> 2015-02-03 17:12 GMT+01:00 Daniel Roe <daniel.r.roe.gmail.com>:
>>> >>>
>>> >>>> Hi,
>>> >>>>
>>> >>>> Not sure this is exactly what you're looking for, but you might want
>>> >>>> to check out the 'watershell' command, which will record the number
>>> of
>>> >>>> water molecules in the 1st and 2nd solvation shells defined by a
>>> lower
>>> >>>> cutoff (default 3.4 Ang.) and an upper cutoff (5.0 Ang). This will
>>> >>>> produce a file containing Frame number, # lower, and # upper. You
>>> >>>> could then calculate a histogram which shows where each value peaks
>>> >>>> and use a second run with the 'closest' command to generate
>>> >>>> trajectories containing the first and second or just the first
>>> >>>> solvation shells for further calculation/visualization etc. For
>>> >>>> example:
>>> >>>>
>>> >>>> parm myparm.parm7
>>> >>>> trajin mycrd.nc
>>> >>>> watershell :MOL W1.dat W1 lower 3.0 upper 5.0
>>> >>>> multihist W1[*] out W1.hist.agr bins 50
>>> >>>>
>>> >>>> Hope this helps,
>>> >>>>
>>> >>>> -Dan
>>> >>>>
>>> >>>> On Tue, Feb 3, 2015 at 8:54 AM, James Starlight <
>>> jmsstarlight.gmail.com>
>>> >>>> wrote:
>>> >>>> > Dear Amber users!
>>> >>>> >
>>> >>>> > I'm looking for the possibility to calculate number of the water
>>> >>>> molecules
>>> >>>> > detected within the protein (here I define area of interests as
>>> :MOL)
>>> >>>> > during molecular dynamic simulation. As the result it will be
>>> better to
>>> >>>> > obtain some graph as well showing dependence of the water within 3
>>> A
>>> >>>> of Mol
>>> >>>> > on Y and time on X. Some time ago I did it using some vmd script
>>> >>>> which was
>>> >>>> > lost :) so not I'm looking for another possibility to do this kind
>>> of
>>> >>>> > analysis using cpptraj.
>>> >>>> >
>>> >>>> > I'd be thankful for any help,
>>> >>>> >
>>> >>>> > James
>>> >>>> > _______________________________________________
>>> >>>> > AMBER mailing list
>>> >>>> > AMBER.ambermd.org
>>> >>>> > http://lists.ambermd.org/mailman/listinfo/amber
>>> >>>>
>>> >>>>
>>> >>>>
>>> >>>> --
>>> >>>> -------------------------
>>> >>>> Daniel R. Roe, PhD
>>> >>>> Department of Medicinal Chemistry
>>> >>>> University of Utah
>>> >>>> 30 South 2000 East, Room 307
>>> >>>> Salt Lake City, UT 84112-5820
>>> >>>> http://home.chpc.utah.edu/~cheatham/
>>> >>>> (801) 587-9652
>>> >>>> (801) 585-6208 (Fax)
>>> >>>>
>>> >>>> _______________________________________________
>>> >>>> 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
>>>
>>>
>>>
>>> --
>>> -------------------------
>>> Daniel R. Roe, PhD
>>> Department of Medicinal Chemistry
>>> University of Utah
>>> 30 South 2000 East, Room 307
>>> Salt Lake City, UT 84112-5820
>>> http://home.chpc.utah.edu/~cheatham/
>>> (801) 587-9652
>>> (801) 585-6208 (Fax)
>>>
>>> _______________________________________________
>>> 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



-- 
-------------------------
Daniel R. Roe, PhD
Department of Medicinal Chemistry
University of Utah
30 South 2000 East, Room 307
Salt Lake City, UT 84112-5820
http://home.chpc.utah.edu/~cheatham/
(801) 587-9652
(801) 585-6208 (Fax)
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Feb 06 2015 - 10:00:06 PST
Custom Search