Error with D-amino acids mid-sequence using NAMD 2.12/2.13b2 and CHARMm36

From: Travis Hesketh (travis.hesketh_at_strath.ac.uk)
Date: Mon Nov 05 2018 - 10:04:49 CST

Hi all,

I am fairly new to NAMD, so this could possibly be due to a mistake on my part, but I am running into the following error when trying to simulate proteins with D-amino acids mid sequence using NAMD 2.12 (also occurs in 2.13b2) and CHARMm 36 (the latest version, toppar_c36_jul18.tgz<http://mackerell.umaryland.edu/download.php?filename=CHARMM_ff_params_files/toppar_c36_jul18.tgz> from http://mackerell.umaryland.edu/charmm_ff.shtml#charmm):

FATAL ERROR: UNABLE TO FIND DIHEDRAL PARAMETERS FOR CTD1 C NH1 CT1

I put together some small tripeptide examples to test what was happening. The error occurs with AdAA and dAdAA. It does not occur with AdAdA, dAAA, dAAdA or dAdAdA. In other words, D-amino acids must be mid sequence (not at the N-terminus) and followed by an L-amino acid.

The dihedral CTD1 C NH1 CT1 refers to atoms 15, 21, 26 and 28 in the first two cases, which are the following lines from the AdAA PDB file:
ATOM 15 CA DALA 2 0.015 0.904 -0.100 1.00 0.00 A
ATOM 21 C DALA 2 1.063 -0.122 -0.200 1.00 0.00 A
ATOM 26 N ALA 3 2.418 0.231 -0.100 1.00 0.00 A
ATOM 28 CA ALA 3 3.510 -0.715 -0.200 1.00 0.00 A

The parameter files are loaded in the following order:
parameters /users/travis/.charmm/par_all36_prot.prm
parameters /users/travis/.charmm/par_all36_carb.prm
parameters /users/travis/.charmm/par_all36_lipid.prm
parameters /users/travis/.charmm/par_all35_ethers.prm
parameters /users/travis/.charmm/par_all36_na.prm
parameters /users/travis/.charmm/par_all36_cgenff.prm
parameters /users/travis/.charmm/toppar_water_ions_namd.str
parameters /users/travis/.charmm/toppar_all36_prot_c36_d_aminoacids.str

The conf file, full log and the unsolvated PDB and PSF files can be found in the following pastes:
https://pastebin.com/j7f8uxyJ - conf file
https://pastebin.com/qf4gcYcz - log file
https://pastebin.com/DZ7EaWN3 - pdb
https://pastebin.com/HxP4qkSe - psf

The PSF and PDB files were generated using psfgen, top_all36_prot.rtf and toppar_all36_prot_c36_d_aminoacids.str.

toppar_all36_prot_c36_d_aminoacids.str was modified to remove the CHARMm commands at the beginning of the file, noted below, but is otherwise as supplied:
set nat ?NATC
set app
if "@NAT" ne "?NATC" if @nat ne 0 set app append

When researching the error message, I couldn’t find any NAMD-specific complaints, but there was mention of a similar error being encountered in GROMACS, though there didn’t appear to be a solution: https://mailman-1.sys.kth.se/pipermail/gromacs.org_gmx-users/2018-March/119045.html

If anyone could shed any light on this, it would be much appreciated.

Kind Regards,
Travis Hesketh
PhD student
University of Strathclyde

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2018 - 23:21:30 CST