AMBER Archive (2009)

Subject: RE: [AMBER] database

From: Steve Seibold (
Date: Thu Sep 17 2009 - 08:43:25 CDT

Hi Fransois
I have been attempting to load the pdb file which, as you can see in the attachment is identical in residues, atom names to the mol2 file and still I am getting the error in leap saying it does not have atom types for my pdb. I have also attempted with "TER" and in its absence to no success.

I hope I am not doing something dumb and wasting your time. I am sending you everything I am using. Can you give it a glance and tell me if you have any ideas?
Is the problem that these mol2 files were not made with gaff.dat/antechamber?
Is there a way to convert mol2 files to prepi files if necessary or would this even help?

Thanks, Steve


-----Original Message-----
From: [] On Behalf Of FyD
Sent: Wednesday, September 16, 2009 2:56 PM
To: AMBER Mailing List
Subject: RE: [AMBER] database


> What I am trying to do is build one topology file and one inpcrd
> file containing both ligand (dCTP) and protein.

It is what I guessed ;-)

> I believe I then need to load leaprc.ff03 followed by dCTP2.mol2
> whose atom names have to match those in my protein/ligand PDB file
> (crystal structure)...right?


> The problem I ran into was that using the dCTP2.mol2 atom names
> which are not unique was not working.

Indeed, an atom name has to be unique in a _same_ residue, yes; but an
atom name can be repeated in different residues.

> That is, just calling all phosphates "P" in my protein/ligand PDB
> file (making it the same names as those of the dCTP2.mol2 file)
> fails to work.

P is indeed found three times in dCTP2 but it is "unique" since it is
found in three different residue numbers: i.e. in the numbers 1, 2 &
3. the fact that residues 2 & 3 have the same name does not matter
since they have a different residue number.

In fact, it is exactly the same case than the CA alpha atom in the
ALA-GLY-GLY peptidic sequence.

> So, I guess I need to rename atom names in dCTP2.mol2 and make them
> correspond to those *unique* atom names in my ligand(dCTP)/protein
> file..right?

No I do not think so. This is difficult to say without doing it.
But, I would say that you need to load in LEaP a protein/dCTP2 complex
(in the PDB file format). I would first try the following residue
ATOM protein whatever set of Cartesian coordinates
ATOM etc... ...
ATOM 1 OR P2M x whatever set of Cartesian coordinates
ATOM 2 P P2M x ...
ATOM 3 OP1 P2M x ...
ATOM 4 OP2 P2M x ...
ATOM 5 OP3 P2M x ...
ATOM 6 OR P1 y ...
ATOM 7 P P1 y ...
ATOM 8 OP1 P1 y ...
ATOM 9 OP2 P1 y ...
ATOM 10 OR P1 y ...
ATOM 11 P P1 y ...
ATOM 12 OP1 P1 y ...
ATOM 13 OP2 P1 y ...
ATOM 14 O3' C5 z ...
                  etc for residue 4 until the end of dCTP2

You need to try but I would say that the presence of the TER words
should match the residue number of the mol2 file...

Do not hesitate to ask for more questions her...

regards, Francois

AMBER mailing list

  • application/octet-stream attachment:

AMBER mailing list