On Mon, 2014-09-08 at 13:26 +0530, MOHD HOMAIDUR RAHMAN wrote:
> Dear Jason, Daniel and All Amber user's
>
> Thank you for your help and guidance
>
> I updated AmberTool 13 to AmberTool !4 and I try to understand how code is
> working.
> For this I check with few combination, but the result I got is not clear to
> me.
> I calculated Interaction Energy by pairwise command in cpptraj 14 like
>
> Input file
>
> trajin 6_md1_ureatest.crd
> reference 6_md1_ureatest.rst
> autoimage
> pairwise uu :MOL out pair_uu.txt cuteelec 12.0 cutevdw 12.0
>
> and in other seperately
>
> trajin ura_leap.crd
> reference ura_leap.crd
> pairwise uu :MOL out pair_uu.txt cuteelec 12.0 cutevdw 12.0
I think you are misinterpreting what cuteelec and cutevdw mean. Those
keywords are _not_ distance-based cutoffs. They are energy-based
cutoffs. From the manual:
[cuteelec <cut>] If comparing to a reference, only report delta EELEC > |<cut>|
(default 1.0 kcal/mol).
[cutevdw <cutv>] If comparing to a reference, only report delta EVDW > |<cutv>|
(default 1.0 kcal/mole).
(I made the same incorrect assumption you did, for what it's worth).
The default values are 1 kcal/mol. There is no distance-based cutoff
for the "pairwise" interaction. What it computes is every interaction
between all pairs of atoms specified in your mask that are not excluded
from one another in the exclusion list... both intramolecular and
intermolecular.
>
> Case 1
> I given input for trajin and reference both leap generated crd file and
> with (electrostatic, vander waal cutoff 12.0) or without cutoff and I am
> expecting zero kcal/mol interaction energy. But I got very high value and
> the value is not effected by cutoff distance.
cuteelec and cutevdw are not distance-based cutoffs, so they will have
no effect on the computed energies. And why would you expect 0 energy?
That would only happen if every atom inside :MOL was excluded from every
other atom inside :MOL by virtue of being 1-2, 1-3, or 1-4 partners.
>
> #Frame uu_cut[EVDW] uu_cut[EELEC] uu_ncut[EVDW]
> uu_ncut[EELEC]
> 1 45847167.8970 29301.6094 45847167.8970
> 29301.6094
>
> The same value I am also getting when I am using the leap generated crd and
> md gerated rst file.
>
> Case 2
> I given input to calculate Interaction energy between two pairs atom. For
> this I taken urea oxygen and all urea hydrogen. Again in this case I am
> getting high positive value. But I am expecting some negative value because
> oxygen have negative charge (-0.641711) and hydrogen have positive charge
> (0.338638). In this case I used reference file as leap generated crd file.
Don't forget this also includes all urea hydrogen-hydrogen interactions.
Those will be repulsive. I suspect that is why you are getting positive
electrostatic contributions here (although your numbers appear
unreasonably large -- do the structures look OK when you visualize
them?)
>
> #Frame uu_cut[EVDW] uu_cut[EELEC]
> 1 16926599.0449 1284114.2386
>
> Case 3
> In this case I just used residue wise pairwise Interaction Energy
> calculation. For reference input, I used both leap generated crd and md
> generated restart file. Both cases giving same result and the result is
> given below.
>
> #Frame uu_cut[EVDW] uu_cut[EELEC] uu_ncut[EVDW]
> uu_ncut[EELEC]
> 1 -2711.7978 24536.6033 -2711.7978
> 24536.6033
>
> All this result not making sense to me, So I need guidance and help from
> you all expert.
>
> Other query how pairwise deal with 1-4 electrostatic and this value how
> contributing in total interaction energy. Because my mdout show total EELEC
> positive and and 1-4 EEL negative and this value also not making sense to
> me.
>From everything that I can tell by looking at the code, the 1-4
electrostatic and van der Waals interactions are completely excluded
(that is, they're not included in any energy contribution that is
printed above). I know the 1-4 pairs are in the exclusion list, and
there seems to be no code handling 1-4 interactions specifically (so I
conclude that they must be omitted).
HTH,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Sep 08 2014 - 05:00:02 PDT
