AMBER Archive (2009)

Subject: [AMBER] parameterization of protein-ligand complex

From: cc cc (
Date: Wed Sep 09 2009 - 11:29:07 CDT


My objective is to conduct a production MD run on a protein-ligand complex and thereafter use the single trajectory obtained to calculate delta G using MMPBSA.

However, I am experiencing some problems regarding parameterization of the complex structure.
Here are the steps which I have tried:
1) Docked an organic molecule (31 atoms) into the active site of the PDB structure (240 amino acids).
2) Used the command:$AMBERHOME/exe/antechamber -i myligand.pdb -fi pdb -o -fo gcrt
3) Ran Gaussian03, using, on the undocked version of the organic molecule and the output = myligand.out
4) Next, I ran:
$AMBERHOME/exe/antechamber -i myligand.out -fi gout -o myligand.prepi -fo prepi -c resp -s 2
5) Verified that all the parameters that I required were available.$AMBERHOME/exe/parmchk -i myligand.prepi -f prepi -o myligand.frcmod
6) Ran tLeap on the ligand
source leaprc.gaff
MyLigand = loadamberprep myligand.prepi
loadamberparams myligand.frcmod
check MyLigand
saveamberparm MyLigand MyLigand.crd
solvateoct MyLigand TIP3PBOX 10
saveamberparm MyLigand MyLigand_solvated.crd
7) Ran tleap on the protein alone
source leaprc.ff99SB
prot=loadpdb myprotein.pdb
saveamberparm prot myprotein.crd
solvateoct myprotein TIP3PBOX 10
saveamberparm myprotein myprotein_solvated.crd
8) $AMBERHOME/exe/ambpdb -p < myprotein_solvated.crd> myprot.pdb

9) $AMBERHOME/exe/ambpdb -p < MyLigand_solvated.crd> mylig.pdb
10) cat myprot.pdb mylig.pdb> complex.pdb
11) Run tleap:
source leaprc.ff99SB
source leaprc.gaff
loadamberparams myligand.frcmod
mylig=loadpdb mylig.pdb
complex=loadpdb complex.pdb
saveamberparm complex complex.crd
12) However, this is the output:
FATAL: Atom .R.A does not have a type.
FATAL: Atom .R.A does not have a type.
FATAL: Atom .R.A does not have a type.
FATAL: Atom .R.A does not have a type.
Failed to generate parameters
Parameter file was not saved.

I thought I have already included the missing parameters in the myligand.frcmod file which was indirectly derived from Gaussian03 calculations.
The other concern I have regarding the above-mentioned procedures is that the ligand was eventually located very far away from the protein in the complex.pdb file. Please advise me about how to make use of the spatial arrangement of the protein and the ligand from my docking.

Thank you very much.

Jeff Yeo
Graduate student
National University of Singapore

Drag n’ drop—Get easy photo sharing with Windows Live™ Photos.
AMBER mailing list