AMBER Archive (2000)

Subject: igb=1 and atom ordering

From: Jarrod Smith (jsmith_at_chazin.structbio.vanderbilt.edu)
Date: Tue Apr 25 2000 - 17:54:55 CDT

I posted a question awhile back about AMBER6 choking on my ligand
residue when I set igb=1. I got several responses which pointed out the
fact that there is an assumed proton ordering in the GB code. i.e. H
atoms are assumed to be bonded to the preceding heavy atom, no matter
what is in prmtop. I thought I'd share my workaround to this problem
since it turned out to be simple (and hopefully not completely
obvious!). With the savepdb command, I wrote a pdb file. My previous
ordering had the atoms grouped by element type with the protons at the
end:

C1
C2
C3
N4
O5
H1
H31
H32
H4
H5
...etc.

I edited the pdb file with vi to meet the criterion set by the GB code:

C1
H1
C2
C3
H31
H32
N4
H4
O5
H5
..etc.

Then I started LEaP and read the edited pdb file (*without* my ligand
library loaded) which caused a new residue to be created. The atoms
were added to the new residue with the pdb ordering. Then I
used bondbydistance to define the bonds, and used saveoff to save my
"new" residue.

Now everything seems to work fine. I hope someone finds it useful. 

-- 
Jarrod A. Smith, Ph.D.
Assistant Director, Center for Structural Biology
Molecular Graphics and Computation
Vanderbilt University

jsmith_at_structbio.vanderbilt.edu