>> Just to clarify what I did, I created 2 prmtop files. In one case (the way
>> my original prmtop files were created), the AG ions are put at the end of
>> proteins. Following your notations, it looks as follows:
>>
>> DNA Strand 1
>> TER
>> DNA Strand 2
>> TER
>> Protein chain1
>> TER
>> Protein chain2
>> TER
>> AG1
>> TER
>> AG2
>> TER
>> END
>>
>> Bond command was used to connect AG1 with four residues in Protein Chain
>> 1, and AG2 with four residues in Protein Chain 2. This system was then
>> neutralized with Na+ ions and solvated with WAT molecules. K+ and Cl- were
>> included later to make it look like in "physiological conditions".
>> prmtop1/inpcrd1 were created at the end.
>>
>> The second prmtop was created for the same system (with modified
>> AG position in the .pdb file) with the following sequence:
>>
>> DNA Strand 1
>> TER
>> DNA Strand 2
>> TER
>> Protein chain1
>> TER
>> AG1
>> TER
>> Protein chain2
>> TER
>> AG2
>> TER
>> END
>>
>> Again, bond command was used to connect AG1 with four residues in Protein
>> Chain 1, and AG2 with four residues in Protein Chain 2. Again, the system
>> was solvated with Na+, K+, Cl-, and Water the same way as before.
>> Prmtop2/inpcrd2 were created at the end.
>>
>> When I look at the ATOMS_PER_MOLECULE sections of both prmtop files, I see
>> the following (these sections are identical, but here is the first few
>> lines):
>>
>> ------------- ATOMS_PER_MOLECULE sections in prmtop1 and prmtop2
>> -------------
>>
>> %FLAG ATOMS_PER_MOLECULE
>> %FORMAT(10I8)
>> 725 731 2057 2057 1 1 1 1 1
>> 1
>> 1 1 1 1 1 1 1 1 1
>> 1
>> 1 1 1 1 1 1 1 1 1
>> 1
>> 1 1 1 1 1 1 1 1 1
>> 1
>> .
>
> In the first case, what is the order of the atoms in the prmtop file? That
> is, does leap move AG1 up to be next to protein chain1? That would be great,
> but I wasn't sure LEaP would be smart enough to do that.
No, AG1 is placed right after the second protein sequence (and AG2 is
right after AG1).
> Another way to ask a similar question: does parmed complain if you follow
> Jason's instructions?
I followed that and it gave an error message. I did not see that error
when I created prmtop2 following your suggestion.
> Note that the ATOMS_PER_MOLECULE is interpreted this way: the first molecule
> is atoms 1-725; second molecule is atoms 726-1456; third molecule is atoms
> 1457-3513, fourth molecule is atoms 3514-5570; etc. this won't be correct
> unless leap has reordered the atoms so that AG1 is in the range 1457-3513.
I dont see any re-ordering in prmtop file.
> On the other hand, if AG1 appears in the prmtop file (and hence, in the
> ambpdb-created pdb file) after protein chain2, there is a problem, both with
> misplaced TER cards and with the way the pressure is calculated. [As mentioned
> before, the magnitude of the pressure error may be tiny in this off-by-one
> case....]
In the pressure calculations, are the masses or other parts in prmtop used
at all? In runmd.f, the pressure is calculated as
pressure = (Nkt*3.0-centvir-(atomvir-Eimp_virial))/(3.0*volume)
(where Nkt = natom * kT).
(I am not sure what centvir, atomvir, and Eimp_virial are). If natom is
right (which is the case in (bad) prmtop file), is the pressure right too?
> I hope I'm describing this clearly enough that you can understand what might
> be happening. The second way of creating things (above) is the better one to
> use, and I have a fear that the first way is not working....
Thanks for the information Dave. I will create the systems your way from
now on, but want to understand and make sure that the pressure is either
calculated right or wrong.
Ilyas,
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Tue Jul 03 2012 - 14:00:04 PDT
