AMBER Archive (2009)

Subject: Re: [AMBER] number-of-EPS-exceeds-MAXESP

From: FyD (
Date: Sat Jan 10 2009 - 02:26:13 CST

Dear Karol,

> I have a HEME residue to parametrize.
> I used Gaussian 03 to generate MEP around my molecule.
> Then:
> espgen -i G03.log -o sustiva.esp and I got more than 100 000 ESP points-
> ~110 000.
> Finally I got RESP charges for my heme molecule however charge for Fe atom
> was suspiciously low = +0.2.
> I found in that RESP can't
> handle more than 100 000 ESP points so
> I decreased the number of ESP points: I changed IOp(6/33=2,6/41=10,6/42=17)
> to IOp(6/33=2,6/41=10,6/42=12)and this time RESP charge for iron was fine.

This is indeed a case where using IOp(6/33=2,6/41=10,6/42=17) might be
interesting although this is always difficult to know what should be a
"good" charge value in this case... I would strongly suggest you
comparing charge values using IOp(6/33=2) and (6/33=2,6/41=10,6/42=17)
for different molecular orientations using the rigid body reorientation
algo implemented in R.E.D. to look at the stability of those charge
values (a single molecular orientation each time).

You would have two R.E.D. versions: one with (6/33=2) and the other
modified using (6/33=2,6/41=10,6/42=17) (see the bug fixes on the
R.E.D. web site in your cases; I can also send you the last version of
R.E.D.-III.1 if you wish).

> However I still would like to know how I should recompile RESP to overcome
> this 100 000 points limit.

This is just a problem of format definition in resp.f:

C Unit 10 input of ESP's (mandatory)
C natoms,nesp (2i5)

You need to change the format of nesp... I just looked at a resp
version from amber8:
~ line 930-940: you should find:
         do imol= 1, inmol
         read(10,'(2i5)') inat,nesp !!!! 2i5 !!!!
       . 'Reading esp"s for molecule ',imol,
       . 'total number of atoms = ',inat,
       . 'total number of esp points = ',NESP
          WRITE(6,'(/ /,a)')

You only need to change the format...

regards, Francois

AMBER mailing list