APBSmem version 2.1.0 released Aug 30 2021
Modified the way the membrane potentials are applied. Any calculation with applied membrane potential in version 2.1.0 and forward will have different results than earlier versions if the protein has an aqueous transmembrane pore or if there is an aqueous cavity in the protein in the lower solution space which extends above the bottom of the surrounding membrane.
In the past, the way that we applied a membrane potential was to set the applied potential as the boundary condition below the membrane, and additionally, fill in an effective charge density everywhere in the lower solution space below the membrane. This extra charge density has the effect of propagating the potential at the boundary to the membrane. Without it, the potential applied at the boundary would fall off exponentially from the boundary to the membrane. See the original APBSmem paper (Callenberg et al.) for a description and derivation.
The problem with the above method becomes obvious when you apply it to proteins with aqueous pores or cavities which enter the membrane region. The applied effective charge density was previously filled in everywhere in the lower solvent space below the nominal bottom of the membrane, which would cause an artifical drop in the absolute value of the potential in aqueous pores and cavities. It is obvious what is wrong with this setup when considering a symmetry operation: Flipping the protein upside-down while simultaneously flipping the applied membrane potential should be an approximate symmetry operation. (Absolute energies will be shifted because the mean potential shifted, but gating charge calculations which are a based on a difference in energies, should theoretically return the same result in both configurations.) To restore the symmetry in the calculation, we do one of two things:
1) For proteins with pores, we linearly scale the applied effective charge density from the value in the bulk lower solvent space to zero over the height of the pore.
2) For proteins without pores, we fill in any cavities in the protein which extend higher than the bottom of the surrounding membrane with the full effective charge density. This way the entire lower solution space is filled with one constant effective charge density, and the entire upper solution space is filled with another constant (zero) effective charge density.
Made the grid dimension text boxes into drop down boxes containing only valid entries.
Some minor improvements in finding MSMS (though there may still be some issues).
APBSmem version 2.0.5 released - Mar 02 2020
Bug fix for gating charge calculations. Occasionally a gating charge calculation would produce nonsense results, due to inconsequential apbs error messages interfering with normal apbs output. The solution was to simply pipe the stderr and stdout from apbs to different files.
When abps can not be found on the PATH, APBSmem would prompt the user to locate apbs. On MacOS, the selection dialog could not look inside of apps to find the apbs binary. Now if you select the app (such as /Applications/APBS-1.5.app), APBSmem will find the abps binary inside the app bundle.
Warning: APBS 1.4 will not work for gating charge calculations. Use version 1.5 (recommended) or 1.3.
APBSmem version 2.0.4 released - Nov 08 2017
Minor big fix for the DrawMembrane command line tool for adding the membrane to the .dx files.
APBSmem version 2.0.3 released - Aug 10 2017
New menu item lets you select which APBS binary to use.
Fixed incorrect units in gating charge plot. (Units in log files were correct.)
APBSmem version 2.0.2 released - Apr 07 2016
Fixed two ancient bugs in gating charge calculations which
resulted in incorrect slope except for an input membrane potential of 2.0.
The tooltip for membrane potential now indicates the membrane potential
is given in reduced units (units of kT/e), rather than in mV.
Divide the potential in mV by 25.69 to get the reduced potential.
APBSmem version 2.0.1 released - Jun 17 2015
Bug fix release.
- Operations involving pdb2pqr (charge assignment, forcefield
changes and pdb conversions) should work now if apbsmem was
called from outside apbsmem directory as such:
java -jar path/to/apbsmem.jar
APBSmem version 2.0 released - Jan 26 2015
NEW FEATURES
- PDB2PQR has been integrated into APBSmem. This allows the user to
convert a PDB file to PQR file, alter protonation states and change
forcefield parameters from within APBSmem
- Delta-pKa button has been added to calculate the membrane induced
pKa shift of a residue sidechain
- Ligand salvation energy has been added as new calculation type
(ligand must be provided as a mol2 file when converting a pdb to
a pqr file)
- Representations menu allows the user to change the molecule
representation from the default ball-and-stick representation to
new cartoon representations color coded by side chain protonation
state or protein solvation energy
- Nonpolar energy is now calculated with protein solvation and gating
charge calculations (Michael Sanner’s MSMS program must be installed
for this function)
Different NP models can be selected from the NP models menu. All NP
models are based the change in solvent accessible surface area, but
these models allow the user to select whether or not to include atoms
in the head group region as solvent excluded.
- Flooding algorithm to draw the membrane around the protein. By using
a flood fill method, the membrane will not be drawn in regions separate
from the bulk membrane area (such as aqueous pores and cavities). This
feature was recently developed and has some limitations. Ensure that
the final focus volume encompasses the protein in the x-y plane, and
make sure that the grid dimensions and lengths are equal in the x y
and z directions.
REQUIRED SOFTWARE
- APBS 1.4.1, the previous version does not work when the boundary
condition in APBSmem set to "Membrane potential"
- MSMS (optional). Used to calculate solvent accessible surface areas
for non polar energy calculations. APBSmem will just calculate
electrostatic energies if MSMS is not included.
- Python 2.7 (optional). Used to run PDB2PQR for PQR creation/modifications.
- Numpy 1.9 (optional). Used by PDB2PQR to assign parameters to ligands.
KNOWN ISSUES
- Ligand option during pdb setup will fail if APBSmem is opened by double
clicking the icon. Open apbsmem from command line when using ligands.
(java -jar <path-to-apbsmem.jar>)
- MSMS will sometimes fail when calculating non polar energy when the
flooding algorithm is used. This can sometimes be fixed by slightly
changing the grid length for the final focus volume.
- Larger grid dimensions (>129) will cause java to run out of memory
open apbsmem from terminal with extra memory allocation when using
large grid dimensions. (java -Xms512m -Xmx512m -jar apbsmem.jar)
APBSmem version 1.13 released - Nov 06 2013
Version 1.13.
APBSmem can be run in batch mode from the command line now.
In batch mode, the GUI still appears, but the calculation
will start automatically, and APBSmem will exit upon completion.
Thus you can run APBSmem in scripts.
Run normally (or double-click on the jar file):
java -jar apbsmem.jar
Some users may need to give java more memory:
java -Xms400m -Xmx800m -jar apbsmem.jar
Batch mode:
java -jar apbsmem.jar <input-file> <output-directory>
For example, APBSmem comes with the three example calculations
from the APBSmem paper. To run the first example from the command line
in the apbsmem directory:
java -jar apbsmem.jar dist/1_proteinsolvation/1_proteinsolvation.solv.in ../out1
Version 1.12.
Fixes a bug in gating charge calculations introduced in version 1.10.
The membrane potential was not being applied. Please upgrade if
you perform gating charge calculations.
Additionally, during a run, the progress bar should accurately reflect
the amount of work remaining.
Version 1.11.3. Bug fix release.
Fixes potential 1-3 minute hang before the GUI appears.
Version 1.11.2. Bug fix release.
Temporary fix for numerous internationalization bugs:
The locale is forced to US for now. This should prevent
many errors seen in locales which use a comma (,) for
the decimal mark. Correct internationalization support should
be available in a future release.
Version 1.11.1. Bug fix release.
Fixed a bug with version 1.11 which caused incorrect results when
using the "Step Ion" feature.
Version 1.11
The current version adds improved support for protein rotations and
translations, some UI tweaks and some warnings to insure that user
input is sensical.
Version 1.10 added two tools for ion solvation calculations:
Menu item Ion -> Create Ion. After picking a protein (i.e. PQR file 1),
you can use the Create Ion tool to place an ion in the system. This will
allow you to quickly create or edit the location of an ion for PQR 2
without writing your own PQR files by hand.
Menu item Ion -> Step Ion. After specifing a protein for PQR file 1 and
an ion for PQR file 2 (by browsing or by creating a ion with the Create
Ion tool) you can use the Step Ion tool to run a series of ion solvation
calculations, stepping the ion over a number of locations.
Version 1.09 added a number of tools under the Orient menu which can be used
to display and alter the orientation of the protein in PQR file 1:
Toggle display of the coordinate axes of the membrane or the protein.
Rotate and center the protein.
Auto-orient the protein. Auto-orientation is very simplistic and can
only work highly symmetric proteins at the moment. It simply aligns
the (unweighted) principal axes of the molecule to the membrane axes.
