Analyzing a Charmm/NAMD generated trajectory in Amber
To convert a Charmm DCD trajectory into an Amber trajectory
that can be interpreted fully with the Amber suite,
several obstacles have to be overcome.
The difficulty stems from the fact that the information in the
topology and parameter files used in Charmm is contained in
differently organized and formatted files in Amber,
and not every feature in the Charmm force field has an
obvious counterpart in Amber.
The solution for this problem calls for defining the system
independently in Amber (e.g., using Leap) and establish
the correspondence between the atom names in the original system and
in the newly defined systems.
After that, the file format conversion can be executed
in such a way that the atoms are rearranged in the converted trajectory
to the newly established order.
Simulaid can establish the correspondence and make the conversion,
including the rearrangement,
but there are several steps on the way
that the user has to perform 'manually'.
- The first step is to generate the .top and .pdb files in Leap.
Simulaid can provide a list of residues to help in this.
If there are non-standard residues in the system,
then the Amber-formatted residue file has to be generated
for that residue. This can also be done with Leap.
Also, the program Intocham can generate this file from an InsightII .car file.
- The .pdb file written by Leap will be a 'regular' PDB file
(i.e., containing atom names of the form 1HA2 instead of HA12,
and atom names like CA will have a space before them).
This file has to undergo a conversion by Simulaid to
'undo regularization' - an option of atom and residue name conversions.
- Similarly, if the Charmm CRD or PDB file does not have the atom names
left-adjusted, it has to undergo the 'undo regularization' step.
- In addition, the Charmm .pdb or .CRD file has to be modified
to reflect the fact that the terminal groups are considered
separate residues in Amber.
This means manually changing the residue numbers in the
Charmm coordinate file. This can be followed by a cleanup
with Simulaid to keep the residue numbers consecutive.
- Probably the hardest part is making sure that all atoms are
The matching of atom and residue names is controlled by the
file pdb_nam.dat, that is part of the distribution.
When Simulaid runs it looks for this file in the current directory
and only if it does not find it there goes to the distribution directory.
Since this file is likely to be in need of modification/extension,
it should be copied into the working directory directory.
Currently, this file contains the standard amino and nucleic acid residues.
Additional residues can be added to it - its syntax is described
in the Simulaid documentation.
- The conversion itself can be invoked as a file format conversion;
convert to Amber trajectory.
During the quiz it will ask the user if rearrangement to a
target order is required.
Answering yes, Simulaid will ask for a structure file
with the target order - the user should specify the Amber .pdb file
that resulted after the 'undo regularization' operation.
Simulaid will then ask for the name conventions for the
input and target system; specify Charmm and Amber, resp.
Simulaid than proceeds with establishing the atom matches.
It is likely that some atoms - especially around the terminal groups
will not be matched. The list of matches is written on a
separate file where the user can see the atoms
that Simulaid could not match - this information should allow
the modification of pdb_nam.dat to achieve the match.
If all atoms are matched successfully then the conversion
can proceed and the converted trajectory should correspond to the
.top file generated by Leap.