Re: [AMBER] bad atom type: OA

From: Jason Swails <>
Date: Tue, 20 Jan 2015 08:19:59 -0500

On Tue, Jan 20, 2015 at 5:59 AM, Houcemeddine Othman <>

> Dear Amber users,
> I have been trying to calculate the per residue free energy decomposition
> of a protein-protein complex. The receptor protein is a metallaoenzyme in
> which two zinc clusters were modeled using MCPB as described in the MTK++
> documentation (HHHO and HHHD configurations). As far as I use only the PB
> namelist the calculation run appropriately. However when adding the GB
> section the following error message appears:
> Beginning GB calculations with /amber12/bin/sander
> calculating complex contribution...
> bad atom type: OA
> CalcError: /amber12/bin/sander failed with prmtop
> /home/free_energy_calculation/complex_vacum.prmtop!
> Exiting. All files have been retained.
> Free energy estimation however were achieved successfully with either &PB
> or &GB namelists.
> Noting that i have amber12 compiled with AmerTools13 just three days ago (I
> also tried it with AmbrtTools12 and it resulted to the same error).
> >From previous mail threads, it seems that only this one is relevant
> However, I can not follow the same solution as it proposes to upgrade to
> amber14 which is not available for me right now.

​Why not? It's freely available at (for this, all you
need is AmberTools 14, not Amber 14).

> Can somebody help me to fix this problem in amber12 ?

​Look for "bad atom type" in $AMBERHOME/src/sander/mdread.F90 (ca. line
2127). You need to add logic there to handle your new element. One option
you have is to replace the error message and mexit call with a default
radius assignment (I would, rather arbitrarily, choose the value for
Carbon, personally). Then you need to recompile for those changes to take

A related post from a couple days ago:


Jason M. Swails
Rutgers University
Postdoctoral Researcher
AMBER mailing list
Received on Tue Jan 20 2015 - 05:30:02 PST
Custom Search