iAPBS/NAMD Interface: User Guide

1. Introduction

The NAMD/APBS interface allows access to APBS functionality from NAMD. The APBS NAMD module can be used to perform implicit solvent MD simulation with NAMD, to write out electrostatic maps for visualization purposes and also to perform MM-PBSA calculations directly with NAMD.

2. Installing the iAPBS Interface

2.1. Requirements

To compile iAPBS/NAMD interface two packages are required:

APBS (version 1.4 or later).
NAMD (version 2.11 or later)

2.2. Building iAPBS interface

Building of the iAPBS interface requires compilation and installation of MALOC and APBS libraries. The following describes each step separately (assumes bash as login script, modify appropriately for t/csh).

# create a build directory and cd to it, then:
P=`pwd`
export APBS_PREFIX=${P}
export APBS_SRC=${P}/apbs-pdb2pqr/apbs
export MCSH_HOME=/dev/null

# get the source
git clone https://github.com/Electrostatics/apbs-pdb2pqr
cd apbs-pdb2pqr
git submodule init
git submodule update
mkdir apbs/build

# configure and build it
cd apbs/build
cmake -DCMAKE_INSTALL_PREFIX:PATH=${APBS_PREFIX} \
 -DCMAKE_BUILD_TYPE=Release -DBUILD_DOC=OFF \
 -DBUILD_SHARED_LIBS=OFF -DENABLE_QUIET=ON \
 -DENABLE_iAPBS=ON -DBUILD_WRAPPER=ON ..

make install

2.3. Building NAMD/APBS module

Currently the NAMD/APBS module is supported on the x86_64 Linux platform only. The Intel compiler is recommended.

cd $APBS_PREFIX
#untar NAMD 2.11 source here
ln -s NAMD_2.11_Source namd2
export NAMD_SRC=$P/namd2
cd $NAMD_SRC
tar xf charm-6.7.0.tar
cd charm-6.7.0
./build charm++ multicore-linux64 --with-production
cd $NAMD_SRC
wget http://www.ks.uiuc.edu/Research/namd/libraries/fftw-linux-x86_64.tar.gz
tar xzf fftw-linux-x86_64.tar.gz
mv linux-x86_64 fftw
wget http://www.ks.uiuc.edu/Research/namd/libraries/tcl8.5.9-linux-x86_64.tar.gz
wget http://www.ks.uiuc.edu/Research/namd/libraries/tcl8.5.9-linux-x86_64-threaded.tar.gz
tar xzf tcl8.5.9-linux-x86_64.tar.gz
tar xzf tcl8.5.9-linux-x86_64-threaded.tar.gz
mv tcl8.5.9-linux-x86_64 tcl
mv tcl8.5.9-linux-x86_64-threaded tcl-threaded

cd src
ln -s $APBS_PREFIX/apbs-pdb2pqr/apbs/contrib/iapbs/modules/NAMD/GlobalMasterAPBS.* .
cd ../arch
ln -s $APBS_PREFIX/apbs-pdb2pqr/apbs/contrib/iapbs/modules/NAMD/*.apbs .
cd $NAMD_SRC
patch -p0 < $APBS_PREFIX/apbs-pdb2pqr/apbs/contrib/iapbs/modules/NAMD/patches/namd2-apbs-2.11.patch

./config apbs Linux-x86_64-g++ --charm-arch multicore-linux64
cd Linux-x86_64-g++
make

# run tests
cd $APBS_PREFIX/apbs-pdb2pqr/apbs/contrib/iapbs/modules/NAMD/test/dipeptide
$NAMD_SRC/Linux-x86_64-g++/namd2 apbs-solvation.conf

cd ../ubq/
$NAMD_SRC/Linux-x86_64-g++/namd2 ubq.conf

3. Using NAMD/APBS module

The APBS module in NAMD adds implicit solvent simulation environment to the NAMD capabilities. The APBS module can be used for example for implicit solvent minimization and dynamics, calculation and visualization of miscellaneous electrostatic biomolecular properties. Please see the Examples section of this User’s Guide.

The following table lists all NAMD/APBS keywords with their description. The left side of the table also lists corresponding APBS keywords which NAMD/APBS keywords mimic very closely. For detailed discussion of APBS keywords please see APBS documentation.

3.1. NAMD/APBS Module Keywords

Table 1. APBS and NAMDAPBS input parameters
APBS keyword	NAMD/APBS keyword	Description
	apbsForces [off]	Turns on the APBS module.
	apbsPQRFile	PQR file name to be read.
mg-auto/mg-para	calc_type [0]	0: manual MG; 1: autoMG; 2: parallel MG
lpbe/nbpe	nonlin [0]	Linear/full Poisson-Boltzmann equation: 0: linear; 1: non-linear; 4: size-dependent PBE
bcfl	bcfl [1]	Boundary condition method: 0: zero; 1: sdh; 2: mdh; 4: focus
srfm	srfm [1]	Surface calculation method: 0: mol; 1: smol; 2: spl2; 3: spl4
pdie	pdie [2.0]	Solute dielectric
sdie	sdie [78.4]	Solvent dielectric
sdens	sdens [10.0]	Vacc sphere density
srad	srad [1.4]	Solvent radius
swin	swin [0.3]	Cubic spline window
temp	temp [298.15]	Temperature (in K)
gamma	gamma [0.105]	Surface tension for apolar energies/forces (in kJ/mol/A²)
chgm	chgm [1]	Charge discretization method: 0: spl0; 1: spl2; 2: spl4
vol	smvolume [10.0]	The parameter smvolume controls the lattice size (in Angstroms³) used in the SMPBE formalism.
size	smsize [1000.0]	The parameter smsize controls the relative size of the ions (in Angstroms) such that each lattice site can contain a single ion of volume radius³ or size ions of volume radius³/size.
calcenergy	calcenergy [1]	Energy calculation flag: 0: Do not perform energy calculation; 1: Calculate total energy only; 2: Calculate per-atom energy components
calcforce	calcforce [0]	Atomic forces calculation: 0: Do not perform force calculation; 1: Calculate total force only; 2: Calculate per-atom force components
	calcnpenergy [1]	Calculate nonpolar energy [0, 1]. 0: Don’t calculate; 1: SASA-based apolar energy caclulation.
write pot	wpot [off]	Writes electrostatic potential data to iapbs-pot.dx in DX format
write charge	wchg [off]	Writes charge data to iapbs-charge.dx in DX format
write smol	wsmol [off]	Writes molecular surface data to iapbs-smol.dx in DX format
write kappa	wkappa [off]	Writes the ion-accessibility kappa map to iapbs-kappa.dx in DX format
write diel	wdiel [off]	Writes dielectric maps to iapbs-diel[x,y,z].dx in DX format
read charge	rchg [off]	Reads charge data from iapbs-charge.dx in DX format
read kappa	rkappa [off]	Reads the ion-accessibility kappa map from iapbs-kappa.dx in DX format
read diel	rdiel [off]	Reads dielectric maps from iapbs-diel[x,y,z].dx in DX format
ion	ion charge conc radius	Counterion charges (in e), Counterion concentrations (in M), Counterion radii (in A)
dime	dime	Grid dimensions (in x, y and z)
cmeth	cmeth [1]	Centering method: 0: Center on a point 1: Center on a molecule
gcent	center	Grid center if cmeth=0
ccmeth	ccmeth [1]	Coarse grid centering method: 0: Center on a point 1: Center on a molecule
cgcent	ccenter	Coarse grid center if ccmeth=0
fcmeth	fcmeth [1]	Fine grid centering method: 0: Center on a point 1: Center on a molecule
fgcent	fcenter	Fine grid center if fcmeth=0
grid	grid	Grid spacings
glen	glen	Grid side lengths
cglen	cglen	Coarse grid side lengths
fglen	fglen	Fine grid side lengths
pdime	pdime	Grid of processors to be used in parallel calculation
ofrac	ofrac [0.1]	Overlap fraction between processors
	debug [0]	Debugging flag [0-5]
	verbose [0]	Print (verbosity) flag [0-5]
	sp_apbs [off]	Perform a single point (a single electrostatic evaluation) APBS calculation
	recalculateGrid [off]	Recalculate the grid dimensions on the fly.

Notes

Values in square brackets [] are defaults.
When the recalculateGrid and grid keywords are set then the grid size parameters (cglen, fglen and dime) are automatically recalculated on the fly during the simulation. For example if grid is set to [0.5 0.5 0.5] Angstroms the grid size will be automatically adjusted to fit the molecule as the system’s dimensions are changed during the simulation (the requested grid spacing 0.5 A will be maintained). This is the recommended option for most of simulations since this prevents the solute to "escape" the pre-set grid.
To disable creation of io.mc file before attempting any extensive minimization or MD simulation with the APBS module please set the MCSH_HOME environment variable to /dev/null (export MCSH_HOME=/dev/null).

4. Examples of using the NAMD/APBS module

The APBS module is initialized with keyword apbsForces. The APBS calculation set up is specified in the apbsForcesConfig section. Please note that charges and radii definition for the electrostatic calculation must be read from a PQR file. See APBS documentation for PQR file format and description. APBS distribution also contains tools for generating valid PQR files (from a PDB file, for example). The order of atoms in the PQR file must be the same as in the associated .top file. For list of APBS-related keywords see NAMD/APBS keywords table.

4.1. Solvation energy calculation

This is an example of single point solvation energy calculation. The electrostatic calculation is done on 0.5³ A numerical grid. The external charges and radii are read from dipeptide.pqr file. The final solvation energy is printed in the "MISC" energy column (as a sum of electrostatic and non-polar energies).

amber on
parmfile dipeptide.top
ambercoor dipeptide.crd
temperature 300
exclude scaled1-4
1-4scaling 0.8333333

switching on
switchDist 9
cutoff  10
pairListDist 11

outputname output
outputEnergies 1
outputTiming   100
dcdFreq        500
restartFreq    500
wrapWater     on
wrapNearest   on

langevin          on
langevinDamping   2
langevinHydrogen  no
langevinTemp     300

apbsForces       on
apbsPQRFile      dipeptide.pqr
apbsForcesConfig {
  calc_type 0 # mg-manual
  grid 0.5 0.5 0.5
  recalculateGrid on
  srfm 2
  chgm 1
  bcfl 1
  debug 1
  verbose 5
  pdie 2.0000
  sdie 78.5400
  sdens 10.00
  srad 1.40
  swin 0.30
  temp 298.15
  gamma 0.105
  sp_apbs off
  wpot off
}
minimize 0

4.2. Molecular dynamics in implicit solvent using APBS

This calculation performs a MD simulation in implicit solvent (water in this case). The charges and radii definition is read from an external file (dipeptide.pqr) and the electrostatic calculation is done on 33³ points of numerical grid with the grid dimensions specified in the input file.

amber on
parmfile dipeptide.top
ambercoor dipeptide.crd
temperature 300
exclude scaled1-4
1-4scaling 0.8333333

switching on
switchDist 9
cutoff  10
pairListDist 11

outputname output
outputEnergies 1
outputTiming   100
dcdFreq        500
restartFreq    500
wrapWater     on
wrapNearest   on

langevin          off
langevinDamping   2
langevinHydrogen  no
langevinTemp     300

apbsForces       on
apbsPQRFile      dipeptide.pqr
apbsForcesConfig {
  dime 33 33 33
  cglen    17.0071 13.8706 12.3012
  fglen    17.0071 13.8706 12.3012
  srfm 2
  chgm 1
  bcfl 1
  debug 0
  pdie 2.0000
  sdie 78.5400
  sdens 10.00
  srad 1.40
  swin 0.30
  temp 298.15
  gamma 0.105
  sp_apbs off
  wpot off
}
numsteps 100

4.3. Calculation and visualization of electrostatic potential

This calculation writes out the calculated electrostatic potential to a file (iapbs-pot.dx). This potential can be visualized using vmd and pymol. The grid dimensions are 1.0³ A.

amber on
parmfile dipeptide.top
ambercoor dipeptide.crd

temperature 300

exclude scaled1-4
1-4scaling 0.8333333

switching on
switchDist 9
cutoff  10
pairListDist 11

outputname output
outputEnergies 1
outputTiming   100
dcdFreq        500
restartFreq    500
wrapWater     on
wrapNearest   on

langevin          on
langevinDamping   2
langevinHydrogen  no
langevinTemp     300

apbsForces       on
apbsPQRFile      dipeptide.pqr
apbsForcesConfig {
  dime 0 0 0
  grid 1.0 1.0 1.0
  srfm 2
  chgm 1
  bcfl 1
  debug 0
  verbose 2
  pdie 2.0000
  sdie 78.5400
  sdens 10.00
  srad 1.40
  swin 0.30
  temp 298.15
  gamma 0.105
  sp_apbs on
  wpot on
}
minimize 0

Notes

To determine the correct size of the numerical grid enveloping the studied molecule (parameters cglen and fglen) you can use psize.py tool from the APBS distribution.