Hello Amy,
The 3_site.cpin file that you sent does not contain the DELTAGREF pointer necessary to finddgref.py. From the pka_corr line in this file, it seems that the last 3 protonation states that you have are energetically equivalent. In this case, you should modify the STATENE flag in one of the following ways, before providing this cpin file to finddgref.py:
STATENE=0.000000,DELTAGREF,DELTAGREF,DELTAGREF,
or
STATENE=DELTAGREF,0.000000,0.000000,0.000000,
As the relative energies between states are all that matters, finddgref.py should return you the same DELTAGREF value in both cases with a flipped sign (assuming you are using the same random seed in your mdin file(s) ).
You mentioned something about the commas in the cpin file, but I could not see this issue in the files you sent. (perhaphs test.cpin constains further modifications in comparison to the 3_site.cpin file that you sent?)
If your problem persists, please send me the exact finddgref.py command you are using together with all input files so I can reproduce your error here.
I hope this helps,
All the best,
Vinícius Wilian D Cruzeiro
PhD Candidate
Department of Chemistry, Physical Chemistry Division
University of Florida, United States
Voice: +1(352)846-1633
________________________________
From: Amy Rice <arice3.hawk.iit.edu>
Sent: Thursday, November 1, 2018 12:28:54 PM
To: AMBER Mailing List
Subject: [AMBER] CpH simulation- error reading charges in long cpin file
Hi all,
I was attempting to run finddgref.py to calculate the reference energy of a
custom titratable residue. The python script failed with the following
(unhelpful) error message:
------------------------------------------------------
ERROR: The execution of AMBER using the binary
/share/apps/amber/18/bin/pmemd.cuda failed. The execution returned the
following STDERR:
At line 167 of file constantph.F90 (unit = 23, file =
'test.cpin.temporary')
--------------------------------------------------
When I look at the mdout file, there is no additional information, it
merely ends with:
--------------------------------------------------
Implicit solvent radii are ArgH and AspGluO modified Bondi2 radii (mbondi3)
|reading charge increments from file: test.cpin.temporary
-----------------------------------------------------
Next, I tried setting all the reference energies to 0 in the cpin file and
running CpH directly. The simulation again fails at the step where the cpin
file is read in. This time I get a slightly more helpful error message:
----------------------------------------------------
At line 167 of file constantph.F90 (unit = 23, file = 'test.cpin')
Fortran runtime error: Cannot match namelist object name
0.090.09-0.180.090.09-0.180.090.09-0.180.09
-----------------------------------------------------
The mdout file ends at the same place as before:
----------------------------------------------------------
Implicit solvent radii are ArgH and AspGluO modified Bondi2 radii (mbondi3)
|reading charge increments from file: test.cpin
-----------------------------------------------------------
>From the error message above (Cannot match namelist object name
0.090.09-0.180.090.09-0.180.090.09-0.180.09), it seems there is a problem
recognizing some of the commas between atom charges when reading in the
cpin file. I have double checked that the commas are all present in the
original cpin file (included with this email). The group I am titrating is
a phosphate group, with three possible protonation sites (on on each
oxygen). I first tested the residue definition and finddgref.py with only
one protonation site, and was able to calculate the reference energy and
run CpH simulation without error. After running into this problem with the
3-site definition, I went back and tried with just 2 possible protonation
sites. Again, the simulation ran without problem. Finally, I tried with
three protonation states defined, but the second two were just identical
copies of the first one (this was to determine if there was a problem with
the third state I had defined, since 1 and 1/2 ran fine, but 1/2/3 gave the
above error). This simulation also failed with the same error as above
("Fortran runtime error: Cannot match namelist object name
0.090.09-0.180.090.09-0.180.090.09-0.180.09").
I have included the following cpin files with this email:
*3_site.cpin* - the original cpin file with protonation sites on all three
phosphate oxygens (the one I'd ultimately like to use) - 4 states total -
simulation fails
*1_site.cpin* - residue definition with just one oxygen protonation site -
2 states total - simulation runs
*2_site.cpin* - protonation possible on two unique oxygen atoms - 3 states
total - simulation runs
*3_site_copy.cpin* - 3 protonation states but all 3 are identical copies of
the one in "1_state.cpin" - 4 states total - simulation fails
There seems to only be a problem reading in the cpin file with any 3-site
definition, which suggests to me the issue is related to the length of the
cpin file or number of states defined, rather than an error with the cpin
itself. Does anyone have any insight into this problem?
Thank you for the help,
Amy
--
Amy Rice
Ph.D. Student
Physics Department
Illinois Institute of Technology
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Nov 01 2018 - 13:00:02 PDT
