Dear Jason,
thanks for your prompt response.
> Have you tried changing the parameters in the beginning of calcpka
> (e.g.
>
> integer, parameter :: STATEINF_FLD_C = 5
> integer, parameter :: TITR_RES_C = 50
> integer, parameter :: TITR_STATES_C = 200
> integer, parameter :: ATOM_CHRG_C = 1000
>
> ) to accomodate the large number of residues? Did you make any changes to
> dynph.h when you ran the initial constant pH simulations with sander? Any
> changes you made there you will have to duplicate in calcpka.F90.
As we defined exactly 50 titratable groups, I thought the parameter
definitions should be sufficient.
In fact, I actually had changed those definitions to test, whether this
could be the problem, and used larger numbers (factor 4 to 10) to be safe.
Although I noticed, that the definitions here are copied from dynph.h, I
did not make any changes there, because calcpka.F90 seems to be compiled
as a stand-alone executable.
When run with the respective input, it really produced some output, but
that had some strange (i.e. non-printable) characters printed in one
line. This was reproducible.
The reason for that behaviour might be the namelist input of the array,
which is defined explicitely up to the index 50 only. So, when I change
the array boundaries, some elements are not defined any more. I'm not
sure about what the Fortran standard defines as the normal behaviour for
reading an array in such a read-in situation...
I'll have a closer look at that.
Best regards,
Anselm
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Apr 11 2013 - 09:30:03 PDT
