Re: [AMBER] error tleap

From: Gustaf Olsson <>
Date: Tue, 15 Sep 2020 12:27:44 +0000

Hello Myrian

I tried reaching out directly as I suspected that this might not be of “general interest” though I am unsure if that email reached you since there was apparently some problem with my organisations email services during the weekend.

I see an issue with the files you sent. Leap complains about groups including an OS atom. I am unable to find this atom in either of the attached files. This means that either the PDB files attached are not the right ones (the ones creating problems in leap) or there is a problem assigning atom names/types for some reason. If you are still experiencing problems and then if possible, could you send the topology and coordinate files for the structures with problems.

Best regards
// Gustaf

On 11 Sep 2020, at 12:46, MYRIAN TORRES RICO <<>> wrote:

Hi Gustaf,

First of all thank you for your time with me.

I keep trying everything that you tells me, but the pdb files that I put in leap are still exactly the same in my eyes. If it's not much of a hassle, I've attached the .pdb files which I'm working with in tleap, for a second view. The angle / torsion that gives me an error, seems to correspond to Galactosamine, C4 and C6 ... but visually there does not seem to be any wrong link ...

The mdk2_rig.pdb this the problem file, and the md4rig.pdb the correct file. The only difference is the
protein conformation (midkine-2 and midkine-4).

thanks in advance again,


Gustaf Olsson <<>> escribió:

Just to set the stage, I am assuming that “different protein conformation” means structure is different though AA sequence is the same.

   Both of these to examples lists works:
   Same sugar - Different protein conformation
   Different sugar - Same protein conformation

If my assumption is correct, then there should be no problem with parameters for the sugar or the protein as they are covered by the combination:

   Same sugar - Same protein conformation

What’s next?

Do a visual inspection using xLEaP edit mode, VMD or something similar with a GUI. What is the group that “CG - OS - S - O” refers to? This seems to be the last “unknown” group and if you solve this then the problem should be solved.

Is this group present in the working example files (should realistically not be present if they do not produce the same errors). Then what has changed? If this is a product of the export process from docking then correcting the labels to match those in the working examples should solve this issue.

If these atoms are present in the other files it is possible that protonation stages has changed or something regarding their positions/coordinates so that they are interpreted as bonded though should not be. Make sure that the group is interpreted correctly when loaded into leap.

At least a place to start

Best regards
// Gustaf

On 10 Sep 2020, at 11:38, MYRIAN TORRES RICO <<><>> wrote:

Hi Gustaf

I work with complex carbohydrates-protein, in this case I have a
My problem is that the following, to find out which angle / torsion is
the one that is giving me an error, I am launching in the same cluster
(so that the conditions are the same), another compound that has the
same pentasaccharide structure but different conformation of protein
to see the differences in the leap.log and see if there is an error in
the sugar, and by the other hand also another compound that is similar
but different the pentasaccharide structure with the same protein
conformation, to know if the problem comes from the protein. And in
none of these cases does tleap give me this error ... That's why I
don't know what to do anymore ...
Could it be a problem with the docking done earlier?

Thanks in advance,


Gustaf Olsson <<><>> escribió:

I suppose that depends on the definition of “wrong”. This is a long
run with a lot of information so apologies if I might have missed

First things I find are these:
glycamod = loadamberparams
Could not open file
/mnt/netapp1/Store_CSIC/home/csic/car/mtr/myriam/smp-191mkd0syn/parameters/Glycam_06d_mod.dat: No such file

modSUBS = loadamberparams
Loading parameters:
Modified force field files must contain both a MASS and NONB entry,
or neither
Could not load parameter set.
-- no parameters loaded>>

##loadOff ions94.lib
Could not open file
/mnt/netapp1/Store_CSIC/home/csic/car/mtr/myriam/smp-191mdk0syn/parameters/solvents.lib: No such file
Could not open database:

These things seem like they could be a cause for the missing
parameters mentioned later on, if these files contains parameters or
information regarding these particular atoms.

I see some "(UNKNOWN ATOM TYPE: )” showing up, this could be a
potential problem though they do not seem to have anything to do
with the missing parameters later on..

So it seems that whatever you are loading might not get loaded
properly. I assume this is some additional parameters/modifications
and might be a cause for the problems observed below?

  Warning: Close contact of X.XXXXXX angstroms between

Would indicate that there are things that are either uncomfortably
close to each other or something has not been defined correctly so
that “bonded” things have become “unbounded things” that are now way
too close to each other. In the prior case, an energy minimisation
of the structure might help remove this warning.

Checking for bond parameters.
Could not find bond parameter for: OS - CG
Checking for angle parameters.
Could not find angle parameter: OS - CG - H2
There are missing parameters.

This seems self explanatory, either the parameters for the bonds and
angles were not sourced properly or in an appropriate order, or
these are not available and would need to be included from an
external source.

Building bond parameters.
Could not find bond parameter for: OS - CG
Building angle parameters.
Could not find angle parameter: OS - CG - H2
Building proper torsion parameters.
** No torsion terms for OS-CG-OS-CG
Building improper torsion parameters.
** Warning: No sp2 improper torsion term for H-H-N3-H
      atoms are: H1 H2 N H3

I would just probably refer to the statement above.

So potentially you have some things for some atoms defined in some
of the files which are not getting sourced/loaded properly. This
could explain the observed complaints. Or, if these are parameters
for atoms which should be included in a FF, that particular
parameterset is not being sourced or not sourced correctly. I would
start by checking which files are missing and not getting loaded and
if they are important to the warnings presented. In which case
getting them sourced correctly might solve the problem(s).

If that is not the case, make sure that the parameter files sourced
do contain the parameters which are reported “missing”. If they do,
then something is not loaded correctly. If not, you might need to
look into using different parameters.

I would suspect that this is a good place to start anyway.

Best regards
// Gustaf

On 9 Sep 2020, at 13:50, MYRIAN TORRES RICO
<<><><>> wrote:

Building improper torsion parameters.
** Warning: No sp2 improper torsion term for H-H-N3-H
      atoms are: H1 H2 N H3

AMBER mailing list<><>

AMBER mailing list<><>

AMBER mailing list<>

AMBER mailing list<>

AMBER mailing list
Received on Tue Sep 15 2020 - 05:30:02 PDT
Custom Search