Modified dihedral notation

User discussions
1

GROMACS version:
GROMACS modification: No

Hi,
I am working on a set of modified nucleic acids, and I need to define different dihedral parameters for the same dihedral type (e.g., OS-CT-CT-F) but for different RNA residues (rFA, rFC, rFU, rFG). I am planning to follow the notation used for Amber99sb-ILDN ff (“#define torsion_ILE_N_CA_CB_CG2_mult1”).

However, I could not see a reference or details in the manual for using this notation. So, I am curious: is this notation valid? And is there any way to cite this approach if it is used?

Also, I see this is defined under RTP entries and ffbonded.itp. Is there any other section where related details need to be added?

2

I see this notation will be handled by the C preprocessor. It will take some time to test the system; meanwhile, it would be great if developers could suggest whether, with current versions of GROMACS (GROMACS 2023 onwards), this notation would still hold.

3

These are macro definitions use the C-preprocessor syntax. Yes, they still work.

4

Thanks for the clarification!

5

Hi,

I assigned torsions under [dihedrals] section in “rna.rtp” and implemented parameters in ffbonded.itp as shown below.

However, “processed.top” from grompp is missing these entries. Are these macros read by grompp but not displayed in “processed.top”?

I see torsion tags in topol.top, does this confirm grompp is reading these values? Is there any way to print this in processed.top (by adding some lines in source code?)

rna.rtp entry

N1     C1'     C2'     F2'   torsion_UFT_N1_C1p_C2p_F2p_mult1
N1     C1'     C2'     F2'   torsion_UFT_N1_C1p_C2p_F2p_mult2
N1     C1'     C2'     F2'   torsion_UFT_N1_C1p_C2p_F2p_mult3
N1     C1'     C2'     F2'   torsion_UFT_N1_C1p_C2p_F2p_mult4

############

ffbonded.itp entry

#define torsion_UFT_N1_C1p_C2p_F2p_mult1          45.0000     77.826584 1
#define torsion_UFT_N1_C1p_C2p_F2p_mult2          45.0000     37.404960 2
#define torsion_UFT_N1_C1p_C2p_F2p_mult3         315.0000    -20.112488 3
#define torsion_UFT_N1_C1p_C2p_F2p_mult4          90.0000    -17.987016 4

I am updating amber amber14sb_OL15 force field port. For non-terminal RNA residues O3’ connection is indicated as "-O3' P" under [ bondtype ] section. Does this indicate connectivity to next residue?

Thanks in advance,
Mandar

6

The processed.top should be missing the definitions or the macro and have all macro instances replaced by the defined content.

7

Thanks, I see now. In processed.top this macro-related section is empty and the dihedrals are directly assigned based on atom index.

; These should not override the general torsion for the atom types,
; so to avoid introducing extra new atom types we list them explicitly for
; the residues in question:

     ### empty section###

and then under [ dihedrals ] section, entries look like this

  112   113   121   101     9    45.0000     77.826584 1
  112   113   121   101     9    45.0000     37.404960 2
  112   113   121   101     9    315.0000    -20.112488 3
  112   113   121   101     9    90.0000    -17.987016 4