On Mon, Oct 8, 2012 at 6:52 PM, "André C. Stiel" <
andre.stiel.tuebingen.mpg.de> wrote:
> Dear Mailing List...
> I have several problems getting single residue energies out of a system.
> I tried around for a time with various suggestions from the web/mailing
> list, to no avail.
> I have two basic problem:
> 1. I get no vdw and eel
> 2. the RRES/ LRES does not work
> 3. can I do something like GB (i come to understand I maybe need this ;-),
> but with 3-4 internal waters
>
> [snip]
> My output is:
>
> PRINT DECOMP - TOTAL ENERGIES
>
> resid |internal |vdw |eel |pol |sas
> ============================================================
> TDC 1 19.398 0.000 0.000 0.000 0.000
> .....
>
> >From reading I figured out, that this idecomp does not work properly wit
> ntb > 0 and I better should
> strip my molecule of all waters and ions and do igb > 0. Correct?
>
I would say most likely you are correct. My understanding is that the ONLY
parts of the code common to electrostatic energy calculations in PME and GB
is the 1-4 EEL calculation (which is included as part of internal for
idecomp=1). This suggests that the PME code is not set up for
decomposition at all (which is not entirely surprising, given that PME
electrostatic energies are even less pairwise decomposable than GB
electrostatic and solvation free energies).
> So I tried the approach from /test:
>
> #decompose energy on a per-residue basis on the trajectory
> &cntrl
> imin = 5, igb = 5, gbsa = 2,
> ntx = 1, maxcyc = 1,
> ntc = 1, ntf = 3,
> ntb = 0, ntp = 0,
> ntwe = 0, ntpr = 500, ntwx=1,
> cut = 1000.0, idecomp = 1,
> /
> Residues considered as REC
> RRES 1 227
> END
> Residues considered as LIG
> LRES 1 227
> END
> Residues to print
> RES 1 227
> END
> END
>
> With a stripped molecule and .prmtop I get:
>
> FATAL: NATOM mismatch in coord and topology files
> Is there a trick to go round this problem. I tried to produce the new -
> water free - .prmtop with tleap on the original .pdb
> But obviously i doesnt work... but ok, I will figure out this.
>
Try using ParmEd to strip out the waters from your topology file. It
should be able to write out a new topology file without ever having to go
through LEaP. Make sure you use the *exact* same strip mask for ParmEd
that you used for ptraj to generate the 'dry' trajectories.
> Second and Third Problem can be easily explained on this basis. 2. Even
> with the RRES/LRES/RES from the /test file, I got:
>
> LOADING THE DECOMP ATOMS AS GROUPS
>
> ----- READING GROUP 1; TITLE:
> Residues considered as REC
> Number of atoms in this group = 0
> ----- READING GROUP 2; TITLE:
> Residues considered as LIG
> Number of atoms in this group = 0
> ----- READING GROUP 3; TITLE:
> Residues to print
> GRP 3 RES 1 TO 227
> Number of atoms in this group = 3541
>
> So LRES and RRES don´t work. Why is that.
>
I'm not sure. To be honest, I've never understood what this distinction
meant in the decomp code, or what it was actually used for.
> Question 3. Some internal waters are important I guess, can I combine igb
> and a stripped molecule with some internal waters?
>
Yes. Of course, I must include the obligatory comments about how GB was
never intended to deal with actual water, but given the fact that GB decomp
energies are not *really* pairwise decomposable and the results are at best
useful as a qualitative measure for your system, this doesn't seem to be a
serious limitation.
HTH,
Jason
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Oct 08 2012 - 15:00:02 PDT
