I've been able to get through from creating a molecule to running MD in
sander, but it turns out some of my atoms in the initial file are getting
"lost". Somewhere between the conversion from pdb to prepin file, my file
goes from 54 atoms down to 24 atoms. This was noted by D. Case - he
thought I had a 24 atom system, but I initially had a 54 atom system. The
remaining atoms get listed as "Unrecognized atomic name X, exit" and
end up having zeroes for x, y, and z, and charge in the pdb.
My system is a base pair, with no sugars or phosphates, and a multi-ringed
small molecule. I thought I could just add TER cards to the right places
in the pdb file and everything would be okay, but it seems something
complicated is going on. My guess is that it has to do with residue names
and how they are recognized by LEAP or antechamber.
The tutorials that mention changing the pdb file to fit AMBER's needs seem
to be specific to a particular file, and I haven't been able to find a list
of "must do's" for modifying pdb files. Note that I tried both making a
pdb file outside of AMBER, and then using antechamber, and also tried
converting a mol2 file to a pdb file using antechamber, then converting to
prepin.
Maybe I have to do more significant changes since I have essentially three
separate molecules, and a charge on the small molecule. I tried running
the base pair by itself, and half of its atoms were dropped too.
Any ideas or suggestions for references addressing the general issue of
modifying pdbs or the specific issue of multiple molecules and
non-nucleotide residues would help me tremendously.
One quick question also - is connectivity data (CONECT) allowed in AMBER
pdb files?
Thanks,
K. Gilbert
NJIT
Received on Tue Apr 01 2003 - 21:53:00 PST
