Bug #1466

Terminal Residue Charges in amber03.ff

Added by Sean McHugh about 3 years ago. Updated almost 3 years ago.

core library
Target version:
Affected version - extra info:
Affected version:


When comparing the atom charges for Alanine between two different amber forcefields, we find that the charges differ as one would expect.

amber03.ff/aminoacids.rtp:[ ALA ]
amber03.ff/aminoacids.rtp- [ atoms ]
amber03.ff/aminoacids.rtp- N N - 0.404773 1
amber03.ff/aminoacids.rtp- H H 0.294276 2
amber03.ff/aminoacids.rtp- CA CT - 0.027733 3
amber03.ff/aminoacids.rtp- HA H1 0.120802 4

amber99.ff/aminoacids.rtp:[ ALA ]
amber99.ff/aminoacids.rtp- [ atoms ]
amber99.ff/aminoacids.rtp- N N - 0.41570 1
amber99.ff/aminoacids.rtp- H H 0.27190 2
amber99.ff/aminoacids.rtp- CA CT 0.03370 3
amber99.ff/aminoacids.rtp- HA H1 0.08230 4

However, if we compare the atom charges for Alanines at the N-terminus between the same two forcefields, we find that the charges are the same regardless of the amber forcefield used. This is true with every amber forcefield found in the GROMACS package. The same is also true of the C-terminus charges (not shown).

amber03.ff/aminoacids.rtp:[ NALA ]
amber03.ff/aminoacids.rtp- [ atoms ]
amber03.ff/aminoacids.rtp- N N3 0.14140 1
amber03.ff/aminoacids.rtp- H1 H 0.19970 2
amber03.ff/aminoacids.rtp- H2 H 0.19970 3
amber03.ff/aminoacids.rtp- H3 H 0.19970 4

amber99.ff/aminoacids.rtp:[ NALA ]
amber99.ff/aminoacids.rtp- [ atoms ]
amber99.ff/aminoacids.rtp- N N3 0.14140 1
amber99.ff/aminoacids.rtp- H1 H 0.19970 2
amber99.ff/aminoacids.rtp- H2 H 0.19970 3
amber99.ff/aminoacids.rtp- H3 H 0.19970 4

Below is what one would expect the charges to be for NALA in amber03.ff. The charges below were obtained from running the amber03 forcefield in the AMBER package.

1         N3      1    ALA      N      1   - 0.58927  14.010000
2 H 1 ALA H1 2 0.44642 1.008000
3 H 1 ALA H2 3 0.44642 1.008000
4 H 1 ALA H3 4 0.44642 1.008000
5 CT 1 ALA CA 5 0.11387 12.010000
6 HP 1 ALA HA 6 0.06715 1.008000
7 CT 1 ALA CB 7 - 0.20411 12.010000
8 HC 1 ALA HB1 8 0.06306 1.008000
9 HC 1 ALA HB2 9 0.06306 1.008000
10 HC 1 ALA HB3 10 0.06306 1.008000

I found it necessary to bring this to your attention as I thought it may be an issue for other groups attempting to use the amber03.ff as well as other amber forcefields in GROMACS. I am not sure if this correlates with other amino acids at the termini as I have only been using Alanine, but it may be something to look into as well.

Associated revisions

Revision f66763bf (diff)
Added by Rossen Apostolov almost 3 years ago

Updated C-/N-terminal partial charges in Amber03.ff.

At the time of porting the AmberFFs were validated against
AMBER 8 and the results have matched precisely. However,
that specific AMBER version had a bug due to which CT/NT charges in ff03
were in fact using ff94 charges. The bug correspondingly propagated
to the Gromacs ports. In newer versions of AMBER this has been fixed.

The current GROMACS patch uses charges as specified in the
all_aminoct03.lib and all_aminont03.lib files as
taken from the AmberTools14 distribution.

In that distribution (14) seem to be no updates to the ff9x parameters.

Fixes #1466.

Change-Id: Ie6cfea5702500ff6cd5019edb22f224d29135425


#1 Updated by Chris Neale about 3 years ago

I don't think that this is a bug.

Amber ff99SB-ILDN appears to be properly implemented for ALA, GLU, and CYS (I didn't check the others) for N-terminal, C-terminal, and internal placement. Comparing Amber03 to Amber99, the charges differ between ff versions only for internal residue placement (I checked ALA, GLU, CYS).

The Amber03 publication ( ) doesn't mention termini at all. However, the abstract explicitly implies that terminal residues in Amber 03 maintain the same charges as previous version of the Amber ff:

The main-chain torsion parameters were obtained by fitting to the energy profiles of Ace-Ala-Nme and Ace-Gly-Nme di-peptides 
calculated using MP2/cc-pVTZ//HF/6-31G** quantum mechanical methods. All other parameters were taken from the existing AMBER data base. 

Therefore, I think that the Amber 03 implementation in gromacs is faithful to the text of the original paper. I can not comment on the current implementation of Amber 03 in the Amber MD software.

Nevertheless, I do think it's worth coordinating with Amber developers so that this difference is either resolved or made apparent to the user.

#2 Updated by Rossen Apostolov about 3 years ago

  • Status changed from New to Accepted

I have contacted Eric Sorin about the report.

#3 Updated by Rossen Apostolov about 3 years ago

  • Assignee changed from Mark Abraham to Rossen Apostolov

At the time of porting the charges have been compared against results from AMBER 8 and they have matched very precisely the GROMACS ones!

Looking at the FF sources, AMBER 8 has in it's amber8/dat/leap/cmd/leaprc.ff03 definition:

loadOff all_aminoct94.lib
loadOff all_aminont94.lib

This has been updated in later Amber versions. E.g. in the last AMBER 14 amber14/dat/leap/cmd/leaprc.ff03.r1 has:

loadOff all_aminoct03.lib
loadOff all_aminont03.lib

I'll fix the termninal charges. Thanks for reporting!

#4 Updated by Gerrit Code Review Bot about 3 years ago

Gerrit received a related patchset '1' for Issue #1466.
Uploader: Rossen Apostolov ()
Change-Id: Ie6cfea5702500ff6cd5019edb22f224d29135425
Gerrit URL:

#5 Updated by Rossen Apostolov about 3 years ago

  • Status changed from Accepted to Fix uploaded

#6 Updated by Rossen Apostolov almost 3 years ago

  • Status changed from Fix uploaded to Resolved
  • % Done changed from 0 to 100

#7 Updated by Erik Lindahl almost 3 years ago

  • Status changed from Resolved to Closed

Also available in: Atom PDF