AMBER Archive (2004)

Subject: Re: AMBER: pbsa error in mm_pbsa amber 8

From: david.evans_at_ulsop.ac.uk
Date: Tue Dec 14 2004 - 09:28:24 CST

Dear Ben,

Look at
http://amber.scripps.edu/doc8/errata.html

fillratio can be changed using the pb namelist (though it's not in
the manual, I think)
To run it via mm_pbsa.pl you need to edit the create_pbsa_input
routine in mm_pbsa_createinput.pm,
just to add fillratio = 4.0 (or whatever) to the namelist.

Dave

 
---- Message from Ben Cossins <bcc_at_soton.ac.uk> at 2004-12-14
13:41:08 ------
>Hi Dave and Ray,
> Sorry for the delayed reply and thanks for your help. I am still
having
>trouble. My ligands are not so small they are all over 85 atoms. I
have
>increased the fillratio (by this a take it you mean the number of
grid
>points/area) again and am still getting fluctuating solvation
energies.
>However they seem to fluctuate less as I increase the SCALE
parameter. With
>6 grid points per angstrom it fluctates from about -2600 to -50
while the
>surface area values stay stable around 980. I am unable to work
through the
>pbsa code and find what the problem is. Can you give me an idea of
what i
>could try.
>
>sorry to bother you again
>
>Ben
>
>On Friday 10 Dec 2004 21:36, Ray Luo wrote:
>> Hi Dave and Ben,
>>
>> This correct. For small ligands, the default fillratio is too
small and
>> this causes pbsa generate garbage energies. This could be the
reason the
>> solvation energies fluctuate enormously.
>>
>> Best,
>> Ray
>>
>> david.evans_at_ulsop.ac.uk wrote:
>> >The comments in the code suggest that this error might occur when
the
>> >grid spacing is large.
>> >So you could try increasing the fillratio. But I have no detailed
>> >understanding of this program, and the developers will know
better.
>> >
>> >Dave Evans
>> >
>> >---
>> >School of Pharmacy
>> >University of London
>> >
>> >---- Message from Ben Cossins <bcc_at_soton.ac.uk> at 2004-12-09
>> >15:19:20 ------
>> >
>> >>Hey again,
>> >> Sorry I should have said that the calculation does not fail but
>> >
>> >the ligand
>> >
>> >>solvation energies fluctuate enourmousely and have SD's of around
>> >
>> >1900.
>> >
>> >>Thanks
>> >>Ben
>> >>
>> >>On Thursday 09 Dec 2004 14:58, Ben Cossins wrote:
>> >>>Hey,
>> >>> I have been using mm_p-bsa from amber 8 to calculate binding
>> >
>> >energies of
>> >
>> >>>ligand-receptor systems. I get an error from pbsa when it is
>> >
>> >calculating
>> >
>> >>>ligand solvation energy for some ligands and not for others.
>> >>>The message is:-
>> >>>
>> >>>PB Warning in epsbnd(): No neighbor found for reentry boundary
grid
>> >>>
>> >>>I have applied all bug fixes.
>> >>>I have had a look at the code giving this error but dont
instantly
>> >
>> >see what
>> >
>> >>>the error is for. I have seen that there has been a bug fix
(bug
>> >
>> >fix4) a
>> >
>> >>>few lines above this.
>> >>>
>> >>>Can anyone offer me any advice.
>> >>>
>> >>>Thanks
>> >>>
>> >>>Ben
>> >>
>> >>--
>> >>Benjamin Cossins
>> >>Southampton University
>> >>J. W. Essex group.
>>
>>--------------------------------------------------------------------
-
>> >
>> >--
>> >
>> >>The AMBER Mail Reflector
>> >>To post, send mail to amber_at_scripps.edu
>> >>To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>> >
>>
>--------------------------------------------------------------------- 

--
>> >The AMBER Mail Reflector
>> >To post, send mail to amber_at_scripps.edu
>> >To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>> >
>> >.
>
>-- 
>Benjamin Cossins
>Southampton University
>J. W. Essex group.
>---------------------------------------------------------------------
--
>The AMBER Mail Reflector
>To post, send mail to amber_at_scripps.edu
>To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu