ViewVC Help
View File | Revision Log | Show Annotations | View Changeset | Root Listing
root/group/trunk/mattDisertation/oopse.tex
Revision: 1068
Committed: Wed Feb 25 21:48:44 2004 UTC (20 years, 4 months ago) by mmeineke
Content type: application/x-tex
File size: 59248 byte(s)
Log Message:
halfway through revisions in OOPSE

File Contents

# Content
1 \chapter{\label{chapt:oopse}OOPSE: AN OPEN SOURCE OBJECT-ORIENTED PARALLEL SIMULATION ENGINE FOR MOLECULAR DYNAMICS}
2
3
4
5 %% \begin{abstract}
6 %% We detail the capabilities of a new open-source parallel simulation
7 %% package ({\sc oopse}) that can perform molecular dynamics simulations
8 %% on atom types that are missing from other popular packages. In
9 %% particular, {\sc oopse} is capable of performing orientational
10 %% dynamics on dipolar systems, and it can handle simulations of metallic
11 %% systems using the embedded atom method ({\sc eam}).
12 %% \end{abstract}
13
14 \lstset{language=C,frame=TB,basicstyle=\small,basicstyle=\ttfamily, %
15 xleftmargin=0.5in, xrightmargin=0.5in,captionpos=b, %
16 abovecaptionskip=0.5cm, belowcaptionskip=0.5cm}
17
18 \section{\label{oopseSec:foreword}Foreword}
19
20 In this chapter, I present and detail the capabilities of the open
21 source simulation package {\sc oopse}. It is important to note, that a
22 simulation package of this size and scope would not have been possible
23 without the collaborative efforts of my colleagues: Charles
24 F.~Vardeman II, Teng Lin, Christopher J.~Fennell and J.~Daniel
25 Gezelter. Although my contributions to {\sc oopse} are major,
26 consideration of my work apart from the others would not give a
27 complete description to the package's capabilities. As such, all
28 contributions to {\sc oopse} to date are presented in this chapter.
29
30 Charles Vardeman is responsible for the parallelization of the long
31 range forces in {\sc oopse} (Sec.~\ref{oopseSec:parallelization}) as
32 well as the inclusion of the embedded-atom potential for transition
33 metals (Sec.~\ref{oopseSec:eam}). Teng Lin's contributions include
34 refinement of the periodic boundary conditions
35 (Sec.~\ref{oopseSec:pbc}), the z-constraint method
36 (Sec.~\ref{oopseSec:zcons}), refinement of the property analysis
37 programs (Sec.~\ref{oopseSec:props}), and development in the extended
38 system integrators (Sec.~\ref{oopseSec:noseHooverThermo}). Christopher
39 Fennell worked on the symplectic integrator
40 (Sec.~\ref{oopseSec:integrate}) and the refinement of the {\sc ssd}
41 water model (Sec.~\ref{oopseSec:SSD}). Daniel Gezelter lent his
42 talents in the development of the extended system integrators
43 (Sec.~\ref{oopseSec:noseHooverThermo}) as well as giving general
44 direction and oversight to the entire project. My responsibilities
45 covered the creation and specification of {\sc bass}
46 (Sec.~\ref{oopseSec:IOfiles}), the original development of the single
47 processor version of {\sc oopse}, contributions to the extended state
48 integrators (Sec.~\ref{oopseSec:noseHooverThermo}), the implementation
49 of the Lennard-Jones (Sec.~\ref{sec:LJPot}) and {\sc duff}
50 (Sec.~\ref{oopseSec:DUFF}) force fields, and initial implementation of
51 the property analysis (Sec.~\ref{oopseSec:props}) and system
52 initialization (Sec.~\ref{oopseSec:initCoords}) utility programs. {\sc
53 oopse}, like many other Molecular Dynamics programs, is a work in
54 progress, and will continue to be so for many graduate student
55 lifetimes.
56
57 \section{\label{sec:intro}Introduction}
58
59 When choosing to simulate a chemical system with molecular dynamics,
60 there are a variety of options available. For simple systems, one
61 might consider writing one's own programming code. However, as systems
62 grow larger and more complex, building and maintaining code for the
63 simulations becomes a time consuming task. In such cases it is usually
64 more convenient for a researcher to turn to pre-existing simulation
65 packages. These packages, such as {\sc amber}\cite{pearlman:1995} and
66 {\sc charmm}\cite{Brooks83}, provide powerful tools for researchers to
67 conduct simulations of their systems without spending their time
68 developing a code base to conduct their research. This then frees them
69 to perhaps explore experimental analogues to their models.
70
71 Despite their utility, problems with these packages arise when
72 researchers try to develop techniques or energetic models that the
73 code was not originally designed to simulate. Examples of uncommonly
74 implemented techniques and energetics include; dipole-dipole
75 interactions, rigid body dynamics, and metallic embedded
76 potentials. When faced with these obstacles, a researcher must either
77 develop their own code or license and extend one of the commercial
78 packages. What we have elected to do, is develop a package of
79 simulation code capable of implementing the types of models upon which
80 our research is based.
81
82 In developing {\sc oopse}, we have adhered to the precepts of Open
83 Source development, and are releasing our source code with a
84 permissive license. It is our intent that by doing so, other
85 researchers might benefit from our work, and add their own
86 contributions to the package. The license under which {\sc oopse} is
87 distributed allows any researcher to download and modify the source
88 code for their own use. In this way further development of {\sc oopse}
89 is not limited to only the models of interest to ourselves, but also
90 those of the community of scientists who contribute back to the
91 project.
92
93 We have structured this chapter to first discuss the empirical energy
94 functions that {\sc oopse } implements in
95 Sec.~\ref{oopseSec:empiricalEnergy}. Following that is a discussion of
96 the various input and output files associated with the package
97 (Sec.~\ref{oopseSec:IOfiles}). Sec.~\ref{oopseSec:mechanics}
98 elucidates the various Molecular Dynamics algorithms {\sc oopse}
99 implements in the integration of the Newtonian equations of
100 motion. Basic analysis of the trajectories obtained from the
101 simulation is discussed in Sec.~\ref{oopseSec:props}. Program design
102 considerations are presented in Sec.~\ref{oopseSec:design}. And
103 lastly, Sec.~\ref{oopseSec:conclusion} concludes the chapter.
104
105 \section{\label{oopseSec:empiricalEnergy}The Empirical Energy Functions}
106
107 \subsection{\label{oopseSec:atomsMolecules}Atoms, Molecules and Rigid Bodies}
108
109 The basic unit of an {\sc oopse} simulation is the atom. The
110 parameters describing the atom are generalized to make the atom as
111 flexible a representation as possible. They may represent specific
112 atoms of an element, or be used for collections of atoms such as
113 methyl and carbonyl groups. The atoms are also capable of having
114 directional components associated with them (\emph{e.g.}~permanent
115 dipoles). Charges, permanent dipoles, and Lennard-Jones parameters for
116 a given atom type are set in the force field parameter files.
117
118 \begin{lstlisting}[float,caption={[Specifier for molecules and atoms] A sample specification of an Ar molecule},label=sch:AtmMole]
119 molecule{
120 name = "Ar";
121 nAtoms = 1;
122 atom[0]{
123 type="Ar";
124 position( 0.0, 0.0, 0.0 );
125 }
126 }
127 \end{lstlisting}
128
129
130 Atoms can be collected into secondary structures such as rigid bodies
131 or molecules. The molecule is a way for {\sc oopse} to keep track of
132 the atoms in a simulation in logical manner. Molecular units store the
133 identities of all the atoms and rigid bodies associated with
134 themselves, and are responsible for the evaluation of their own
135 internal interactions (\emph{i.e.}~bonds, bends, and torsions). Scheme
136 \ref{sch:AtmMole} shows how one creates a molecule in a ``model'' or
137 \texttt{.mdl} file. The position of the atoms given in the
138 declaration are relative to the origin of the molecule, and is used
139 when creating a system containing the molecule.
140
141 As stated previously, one of the features that sets {\sc oopse} apart
142 from most of the current molecular simulation packages is the ability
143 to handle rigid body dynamics. Rigid bodies are non-spherical
144 particles or collections of particles that have a constant internal
145 potential and move collectively.\cite{Goldstein01} They are not
146 included in most simulation packages because of the algorithmic
147 complexity involved in propagating orientational degrees of
148 freedom. Until recently, integrators which propagate orientational
149 motion have been much worse than those available for translational
150 motion.
151
152 Moving a rigid body involves determination of both the force and
153 torque applied by the surroundings, which directly affect the
154 translational and rotational motion in turn. In order to accumulate
155 the total force on a rigid body, the external forces and torques must
156 first be calculated for all the internal particles. The total force on
157 the rigid body is simply the sum of these external forces.
158 Accumulation of the total torque on the rigid body is more complex
159 than the force because the torque is applied to the center of mass of
160 the rigid body. The torque on rigid body $i$ is
161 \begin{equation}
162 \boldsymbol{\tau}_i=
163 \sum_{a}\biggl[(\mathbf{r}_{ia}-\mathbf{r}_i)\times \mathbf{f}_{ia}
164 + \boldsymbol{\tau}_{ia}\biggr]
165 \label{eq:torqueAccumulate}
166 \end{equation}
167 where $\boldsymbol{\tau}_i$ and $\mathbf{r}_i$ are the torque on and
168 position of the center of mass respectively, while $\mathbf{f}_{ia}$,
169 $\mathbf{r}_{ia}$, and $\boldsymbol{\tau}_{ia}$ are the force on,
170 position of, and torque on the component particles of the rigid body.
171
172 The summation of the total torque is done in the body fixed axis of
173 each rigid body. In order to move between the space fixed and body
174 fixed coordinate axes, parameters describing the orientation must be
175 maintained for each rigid body. At a minimum, the rotation matrix
176 (\textbf{A}) can be described by the three Euler angles ($\phi,
177 \theta,$ and $\psi$), where the elements of \textbf{A} are composed of
178 trigonometric operations involving $\phi, \theta,$ and
179 $\psi$.\cite{Goldstein01} In order to avoid numerical instabilities
180 inherent in using the Euler angles, the four parameter ``quaternion''
181 scheme is often used. The elements of \textbf{A} can be expressed as
182 arithmetic operations involving the four quaternions ($q_0, q_1, q_2,$
183 and $q_3$).\cite{allen87:csl} Use of quaternions also leads to
184 performance enhancements, particularly for very small
185 systems.\cite{Evans77}
186
187 {\sc oopse} utilizes a relatively new scheme that propagates the
188 entire nine parameter rotation matrix. Further discussion
189 on this choice can be found in Sec.~\ref{oopseSec:integrate}. An example
190 definition of a rigid body can be seen in Scheme
191 \ref{sch:rigidBody}. The positions in the atom definitions are the
192 placements of the atoms relative to the origin of the rigid body,
193 which itself has a position relative to the origin of the molecule.
194
195 \begin{lstlisting}[float,caption={[Defining rigid bodies]A sample definition of a rigid body},label={sch:rigidBody}]
196 molecule{
197 name = "TIP3P_water";
198 nRigidBodies = 1;
199 rigidBody[0]{
200 nAtoms = 3;
201 atom[0]{
202 type = "O_TIP3P";
203 position( 0.0, 0.0, -0.06556 );
204 }
205 atom[1]{
206 type = "H_TIP3P";
207 position( 0.0, 0.75695, 0.52032 );
208 }
209 atom[2]{
210 type = "H_TIP3P";
211 position( 0.0, -0.75695, 0.52032 );
212 }
213 position( 0.0, 0.0, 0.0 );
214 orientation( 0.0, 0.0, 1.0 );
215 }
216 }
217 \end{lstlisting}
218
219 \subsection{\label{sec:LJPot}The Lennard Jones Force Field}
220
221 The most basic force field implemented in {\sc oopse} is the
222 Lennard-Jones force field, which mimics the van der Waals interaction at
223 long distances, and uses an empirical repulsion at short
224 distances. The Lennard-Jones potential is given by:
225 \begin{equation}
226 V_{\text{LJ}}(r_{ij}) =
227 4\epsilon_{ij} \biggl[
228 \biggl(\frac{\sigma_{ij}}{r_{ij}}\biggr)^{12}
229 - \biggl(\frac{\sigma_{ij}}{r_{ij}}\biggr)^{6}
230 \biggr]
231 \label{eq:lennardJonesPot}
232 \end{equation}
233 Where $r_{ij}$ is the distance between particles $i$ and $j$,
234 $\sigma_{ij}$ scales the length of the interaction, and
235 $\epsilon_{ij}$ scales the well depth of the potential. Scheme
236 \ref{sch:LJFF} gives and example \texttt{.bass} file that
237 sets up a system of 108 Ar particles to be simulated using the
238 Lennard-Jones force field.
239
240 \begin{lstlisting}[float,caption={[Invocation of the Lennard-Jones force field] A sample system using the Lennard-Jones force field.},label={sch:LJFF}]
241
242 #include "argon.mdl"
243
244 nComponents = 1;
245 component{
246 type = "Ar";
247 nMol = 108;
248 }
249
250 initialConfig = "./argon.init";
251
252 forceField = "LJ";
253 \end{lstlisting}
254
255 Because this potential is calculated between all pairs, the force
256 evaluation can become computationally expensive for large systems. To
257 keep the pair evaluations to a manageable number, {\sc oopse} employs
258 a cut-off radius.\cite{allen87:csl} The cutoff radius can either be
259 specified in the \texttt{.bass} file, or left as its default value of
260 $2.5\sigma_{ii}$, where $\sigma_{ii}$ is the largest Lennard-Jones
261 length parameter present in the simulation. Truncating the calculation
262 at $r_{\text{cut}}$ introduces a discontinuity into the potential
263 energy and the force. To offset this discontinuity in the potential,
264 the energy value at $r_{\text{cut}}$ is subtracted from the
265 potential. This causes the potential to go to zero smoothly at the
266 cut-off radius, and preserves conservation of energy in integrating
267 the equations of motion.
268
269 Interactions between dissimilar particles requires the generation of
270 cross term parameters for $\sigma$ and $\epsilon$. These are
271 calculated through the Lorentz-Berthelot mixing
272 rules:\cite{allen87:csl}
273 \begin{equation}
274 \sigma_{ij} = \frac{1}{2}[\sigma_{ii} + \sigma_{jj}]
275 \label{eq:sigmaMix}
276 \end{equation}
277 and
278 \begin{equation}
279 \epsilon_{ij} = \sqrt{\epsilon_{ii} \epsilon_{jj}}
280 \label{eq:epsilonMix}
281 \end{equation}
282
283
284
285 \subsection{\label{oopseSec:DUFF}Dipolar Unified-Atom Force Field}
286
287 The dipolar unified-atom force field ({\sc duff}) was developed to
288 simulate lipid bilayers. The simulations require a model capable of
289 forming bilayers, while still being sufficiently computationally
290 efficient to allow large systems ($\sim$100's of phospholipids,
291 $\sim$1000's of waters) to be simulated for long times
292 ($\sim$10's of nanoseconds).
293
294 With this goal in mind, {\sc duff} has no point
295 charges. Charge-neutral distributions were replaced with dipoles,
296 while most atoms and groups of atoms were reduced to Lennard-Jones
297 interaction sites. This simplification cuts the length scale of long
298 range interactions from $\frac{1}{r}$ to $\frac{1}{r^3}$, and allows
299 us to avoid the computationally expensive Ewald sum. Instead, we can
300 use neighbor-lists and cutoff radii for the dipolar interactions, or
301 include a reaction field to mimic larger range interactions.
302
303 As an example, lipid head-groups in {\sc duff} are represented as
304 point dipole interaction sites. By placing a dipole at the head group
305 center of mass, our model mimics the charge separation found in common
306 phospholipids such as phosphatidylcholine.\cite{Cevc87} Additionally,
307 a large Lennard-Jones site is located at the pseudoatom's center of
308 mass. The model is illustrated by the red atom in
309 Fig.~\ref{oopseFig:lipidModel}. The water model we use to complement
310 the dipoles of the lipids is our reparameterization of the soft sticky
311 dipole (SSD) model of Ichiye
312 \emph{et al.}\cite{liu96:new_model}
313
314 \begin{figure}
315 \centering
316 \includegraphics[width=\linewidth]{lipidModel.eps}
317 \caption{A representation of the lipid model. $\phi$ is the torsion angle, $\theta$ %
318 is the bend angle, $\mu$ is the dipole moment of the head group, and n
319 is the chain length.}
320 \label{oopseFig:lipidModel}
321 \end{figure}
322
323 We have used a set of scalable parameters to model the alkyl groups
324 with Lennard-Jones sites. For this, we have borrowed parameters from
325 the TraPPE force field of Siepmann
326 \emph{et al}.\cite{Siepmann1998} TraPPE is a unified-atom
327 representation of n-alkanes, which is parametrized against phase
328 equilibria using Gibbs ensemble Monte Carlo simulation
329 techniques.\cite{Siepmann1998} One of the advantages of TraPPE is that
330 it generalizes the types of atoms in an alkyl chain to keep the number
331 of pseudoatoms to a minimum; the parameters for a unified atom such as
332 $\text{CH}_2$ do not change depending on what species are bonded to
333 it.
334
335 TraPPE also constrains all bonds to be of fixed length. Typically,
336 bond vibrations are the fastest motions in a molecular dynamic
337 simulation. Small time steps between force evaluations must be used to
338 ensure adequate energy conservation in the bond degrees of freedom. By
339 constraining the bond lengths, larger time steps may be used when
340 integrating the equations of motion. A simulation using {\sc duff} is
341 illustrated in Scheme \ref{sch:DUFF}.
342
343 \begin{lstlisting}[float,caption={[Invocation of {\sc duff}]Sample \texttt{.bass} file showing a simulation utilizing {\sc duff}},label={sch:DUFF}]
344
345 #include "water.mdl"
346 #include "lipid.mdl"
347
348 nComponents = 2;
349 component{
350 type = "simpleLipid_16";
351 nMol = 60;
352 }
353
354 component{
355 type = "SSD_water";
356 nMol = 1936;
357 }
358
359 initialConfig = "bilayer.init";
360
361 forceField = "DUFF";
362
363 \end{lstlisting}
364
365 \subsection{\label{oopseSec:energyFunctions}{\sc duff} Energy Functions}
366
367 The total potential energy function in {\sc duff} is
368 \begin{equation}
369 V = \sum^{N}_{I=1} V^{I}_{\text{Internal}}
370 + \sum^{N-1}_{I=1} \sum_{J>I} V^{IJ}_{\text{Cross}}
371 \label{eq:totalPotential}
372 \end{equation}
373 Where $V^{I}_{\text{Internal}}$ is the internal potential of molecule $I$:
374 \begin{equation}
375 V^{I}_{\text{Internal}} =
376 \sum_{\theta_{ijk} \in I} V_{\text{bend}}(\theta_{ijk})
377 + \sum_{\phi_{ijkl} \in I} V_{\text{torsion}}(\phi_{ijkl})
378 + \sum_{i \in I} \sum_{(j>i+4) \in I}
379 \biggl[ V_{\text{LJ}}(r_{ij}) + V_{\text{dipole}}
380 (\mathbf{r}_{ij},\boldsymbol{\Omega}_{i},\boldsymbol{\Omega}_{j})
381 \biggr]
382 \label{eq:internalPotential}
383 \end{equation}
384 Here $V_{\text{bend}}$ is the bend potential for all 1, 3 bonded pairs
385 within the molecule $I$, and $V_{\text{torsion}}$ is the torsion potential
386 for all 1, 4 bonded pairs. The pairwise portions of the internal
387 potential are excluded for pairs that are closer than three bonds,
388 i.e.~atom pairs farther away than a torsion are included in the
389 pair-wise loop.
390
391
392 The bend potential of a molecule is represented by the following function:
393 \begin{equation}
394 V_{\text{bend}}(\theta_{ijk}) = k_{\theta}( \theta_{ijk} - \theta_0 )^2 \label{eq:bendPot}
395 \end{equation}
396 Where $\theta_{ijk}$ is the angle defined by atoms $i$, $j$, and $k$
397 (see Fig.~\ref{oopseFig:lipidModel}), $\theta_0$ is the equilibrium
398 bond angle, and $k_{\theta}$ is the force constant which determines the
399 strength of the harmonic bend. The parameters for $k_{\theta}$ and
400 $\theta_0$ are borrowed from those in TraPPE.\cite{Siepmann1998}
401
402 The torsion potential and parameters are also borrowed from TraPPE. It is
403 of the form:
404 \begin{equation}
405 V_{\text{torsion}}(\phi) = c_1[1 + \cos \phi]
406 + c_2[1 + \cos(2\phi)]
407 + c_3[1 + \cos(3\phi)]
408 \label{eq:origTorsionPot}
409 \end{equation}
410 Where:
411 \begin{equation}
412 \cos\phi = (\hat{\mathbf{r}}_{ij} \times \hat{\mathbf{r}}_{jk}) \cdot
413 (\hat{\mathbf{r}}_{jk} \times \hat{\mathbf{r}}_{kl})
414 \label{eq:torsPhi}
415 \end{equation}
416 Here, $\hat{\mathbf{r}}_{\alpha\beta}$ are the set of unit bond
417 vectors between atoms $i$, $j$, $k$, and $l$. For computational
418 efficiency, the torsion potential has been recast after the method of
419 {\sc charmm},\cite{Brooks83} in which the angle series is converted to
420 a power series of the form:
421 \begin{equation}
422 V_{\text{torsion}}(\phi) =
423 k_3 \cos^3 \phi + k_2 \cos^2 \phi + k_1 \cos \phi + k_0
424 \label{eq:torsionPot}
425 \end{equation}
426 Where:
427 \begin{align*}
428 k_0 &= c_1 + c_3 \\
429 k_1 &= c_1 - 3c_3 \\
430 k_2 &= 2 c_2 \\
431 k_3 &= 4c_3
432 \end{align*}
433 By recasting the potential as a power series, repeated trigonometric
434 evaluations are avoided during the calculation of the potential energy.
435
436
437 The cross potential between molecules $I$ and $J$, $V^{IJ}_{\text{Cross}}$, is
438 as follows:
439 \begin{equation}
440 V^{IJ}_{\text{Cross}} =
441 \sum_{i \in I} \sum_{j \in J}
442 \biggl[ V_{\text{LJ}}(r_{ij}) + V_{\text{dipole}}
443 (\mathbf{r}_{ij},\boldsymbol{\Omega}_{i},\boldsymbol{\Omega}_{j})
444 + V_{\text{sticky}}
445 (\mathbf{r}_{ij},\boldsymbol{\Omega}_{i},\boldsymbol{\Omega}_{j})
446 \biggr]
447 \label{eq:crossPotentail}
448 \end{equation}
449 Where $V_{\text{LJ}}$ is the Lennard Jones potential,
450 $V_{\text{dipole}}$ is the dipole dipole potential, and
451 $V_{\text{sticky}}$ is the sticky potential defined by the SSD model
452 (Sec.~\ref{oopseSec:SSD}). Note that not all atom types include all
453 interactions.
454
455 The dipole-dipole potential has the following form:
456 \begin{equation}
457 V_{\text{dipole}}(\mathbf{r}_{ij},\boldsymbol{\Omega}_{i},
458 \boldsymbol{\Omega}_{j}) = \frac{|\mu_i||\mu_j|}{4\pi\epsilon_{0}r_{ij}^{3}} \biggl[
459 \boldsymbol{\hat{u}}_{i} \cdot \boldsymbol{\hat{u}}_{j}
460 -
461 3(\boldsymbol{\hat{u}}_i \cdot \hat{\mathbf{r}}_{ij}) %
462 (\boldsymbol{\hat{u}}_j \cdot \hat{\mathbf{r}}_{ij}) \biggr]
463 \label{eq:dipolePot}
464 \end{equation}
465 Here $\mathbf{r}_{ij}$ is the vector starting at atom $i$ pointing
466 towards $j$, and $\boldsymbol{\Omega}_i$ and $\boldsymbol{\Omega}_j$
467 are the orientational degrees of freedom for atoms $i$ and $j$
468 respectively. $|\mu_i|$ is the magnitude of the dipole moment of atom
469 $i$, $\boldsymbol{\hat{u}}_i$ is the standard unit orientation vector
470 of $\boldsymbol{\Omega}_i$, and $\boldsymbol{\hat{r}}_{ij}$ is the
471 unit vector pointing along $\mathbf{r}_{ij}$
472 ($\boldsymbol{\hat{r}}_{ij}=\mathbf{r}_{ij}/|\mathbf{r}_{ij}|$).
473
474 To improve computational efficiency of the dipole-dipole interactions,
475 {\sc oopse} employs an electrostatic cutoff radius. This parameter can
476 be set in the \texttt{.bass} file, and controls the length scale over
477 which dipole interactions are felt. To compensate for the
478 discontinuity in the potential and the forces at the cutoff radius, we
479 have implemented a switching function to smoothly scale the
480 dipole-dipole interaction at the cutoff.
481 \begin{equation}
482 S(r_{ij}) =
483 \begin{cases}
484 1 & \text{if $r_{ij} \le r_t$},\\
485 \frac{(r_{\text{cut}} + 2r_{ij} - 3r_t)(r_{\text{cut}} - r_{ij})^2}
486 {(r_{\text{cut}} - r_t)^2}
487 & \text{if $r_t < r_{ij} \le r_{\text{cut}}$}, \\
488 0 & \text{if $r_{ij} > r_{\text{cut}}$.}
489 \end{cases}
490 \label{eq:dipoleSwitching}
491 \end{equation}
492 Here $S(r_{ij})$ scales the potential at a given $r_{ij}$, and $r_t$
493 is the taper radius some given thickness less than the electrostatic
494 cutoff. The switching thickness can be set in the \texttt{.bass} file.
495
496 \subsection{\label{oopseSec:SSD}The {\sc duff} Water Models: SSD/E and SSD/RF}
497
498 In the interest of computational efficiency, the default solvent used
499 by {\sc oopse} is the extended Soft Sticky Dipole (SSD/E) water
500 model.\cite{Gezelter04} The original SSD was developed by Ichiye
501 \emph{et al.}\cite{liu96:new_model} as a modified form of the hard-sphere
502 water model proposed by Bratko, Blum, and
503 Luzar.\cite{Bratko85,Bratko95} It consists of a single point dipole
504 with a Lennard-Jones core and a sticky potential that directs the
505 particles to assume the proper hydrogen bond orientation in the first
506 solvation shell. Thus, the interaction between two SSD water molecules
507 \emph{i} and \emph{j} is given by the potential
508 \begin{equation}
509 V_{ij} =
510 V_{ij}^{LJ} (r_{ij})\ + V_{ij}^{dp}
511 (\mathbf{r}_{ij},\boldsymbol{\Omega}_i,\boldsymbol{\Omega}_j)\ +
512 V_{ij}^{sp}
513 (\mathbf{r}_{ij},\boldsymbol{\Omega}_i,\boldsymbol{\Omega}_j),
514 \label{eq:ssdPot}
515 \end{equation}
516 where the $\mathbf{r}_{ij}$ is the position vector between molecules
517 \emph{i} and \emph{j} with magnitude equal to the distance $r_{ij}$, and
518 $\boldsymbol{\Omega}_i$ and $\boldsymbol{\Omega}_j$ represent the
519 orientations of the respective molecules. The Lennard-Jones and dipole
520 parts of the potential are given by equations \ref{eq:lennardJonesPot}
521 and \ref{eq:dipolePot} respectively. The sticky part is described by
522 the following,
523 \begin{equation}
524 u_{ij}^{sp}(\mathbf{r}_{ij},\boldsymbol{\Omega}_i,\boldsymbol{\Omega}_j)=
525 \frac{\nu_0}{2}[s(r_{ij})w(\mathbf{r}_{ij},
526 \boldsymbol{\Omega}_i,\boldsymbol{\Omega}_j) +
527 s^\prime(r_{ij})w^\prime(\mathbf{r}_{ij},
528 \boldsymbol{\Omega}_i,\boldsymbol{\Omega}_j)]\ ,
529 \label{eq:stickyPot}
530 \end{equation}
531 where $\nu_0$ is a strength parameter for the sticky potential, and
532 $s$ and $s^\prime$ are cubic switching functions which turn off the
533 sticky interaction beyond the first solvation shell. The $w$ function
534 can be thought of as an attractive potential with tetrahedral
535 geometry:
536 \begin{equation}
537 w({\bf r}_{ij},{\bf \Omega}_i,{\bf \Omega}_j)=
538 \sin\theta_{ij}\sin2\theta_{ij}\cos2\phi_{ij},
539 \label{eq:stickyW}
540 \end{equation}
541 while the $w^\prime$ function counters the normal aligned and
542 anti-aligned structures favored by point dipoles:
543 \begin{equation}
544 w^\prime({\bf r}_{ij},{\bf \Omega}_i,{\bf \Omega}_j)=
545 (\cos\theta_{ij}-0.6)^2(\cos\theta_{ij}+0.8)^2-w^0,
546 \label{eq:stickyWprime}
547 \end{equation}
548 It should be noted that $w$ is proportional to the sum of the $Y_3^2$
549 and $Y_3^{-2}$ spherical harmonics (a linear combination which
550 enhances the tetrahedral geometry for hydrogen bonded structures),
551 while $w^\prime$ is a purely empirical function. A more detailed
552 description of the functional parts and variables in this potential
553 can be found in the original SSD
554 articles.\cite{liu96:new_model,liu96:monte_carlo,chandra99:ssd_md,Ichiye03}
555
556 Since SSD/E is a single-point {\it dipolar} model, the force
557 calculations are simplified significantly relative to the standard
558 {\it charged} multi-point models. In the original Monte Carlo
559 simulations using this model, Ichiye {\it et al.} reported that using
560 SSD decreased computer time by a factor of 6-7 compared to other
561 models.\cite{liu96:new_model} What is most impressive is that these savings
562 did not come at the expense of accurate depiction of the liquid state
563 properties. Indeed, SSD/E maintains reasonable agreement with the Head-Gordon
564 diffraction data for the structural features of liquid
565 water.\cite{hura00,liu96:new_model} Additionally, the dynamical properties
566 exhibited by SSD/E agree with experiment better than those of more
567 computationally expensive models (like TIP3P and
568 SPC/E).\cite{chandra99:ssd_md} The combination of speed and accurate depiction
569 of solvent properties makes SSD/E a very attractive model for the
570 simulation of large scale biochemical simulations.
571
572 Recent constant pressure simulations revealed issues in the original
573 SSD model that led to lower than expected densities at all target
574 pressures.\cite{Ichiye03,Gezelter04} The default model in {\sc oopse}
575 is therefore SSD/E, a density corrected derivative of SSD that
576 exhibits improved liquid structure and transport behavior. If the use
577 of a reaction field long-range interaction correction is desired, it
578 is recommended that the parameters be modified to those of the SSD/RF
579 model. Solvent parameters can be easily modified in an accompanying
580 \texttt{.bass} file as illustrated in the scheme below. A table of the
581 parameter values and the drawbacks and benefits of the different
582 density corrected SSD models can be found in
583 reference~\cite{Gezelter04}.
584
585 \begin{lstlisting}[float,caption={[A simulation of {\sc ssd} water]An example file showing a simulation including {\sc ssd} water.},label={sch:ssd}]
586
587 #include "water.mdl"
588
589 nComponents = 1;
590 component{
591 type = "SSD_water";
592 nMol = 864;
593 }
594
595 initialConfig = "liquidWater.init";
596
597 forceField = "DUFF";
598
599 /*
600 * The following two flags set the cutoff
601 * radius for the electrostatic forces
602 * as well as the skin thickness of the switching
603 * function.
604 */
605
606 electrostaticCutoffRadius = 9.2;
607 electrostaticSkinThickness = 1.38;
608
609 \end{lstlisting}
610
611
612 \subsection{\label{oopseSec:eam}Embedded Atom Method}
613
614 There are Molecular Dynamics packages which have the
615 capacity to simulate metallic systems, including some that have
616 parallel computational abilities\cite{plimpton93}. Potentials that
617 describe bonding transition metal
618 systems\cite{Finnis84,Ercolessi88,Chen90,Qi99,Ercolessi02} have an
619 attractive interaction which models ``Embedding''
620 a positively charged metal ion in the electron density due to the
621 free valance ``sea'' of electrons created by the surrounding atoms in
622 the system. A mostly-repulsive pairwise part of the potential
623 describes the interaction of the positively charged metal core ions
624 with one another. A particular potential description called the
625 Embedded Atom Method\cite{Daw84,FBD86,johnson89,Lu97}({\sc eam}) that has
626 particularly wide adoption has been selected for inclusion in {\sc oopse}. A
627 good review of {\sc eam} and other metallic potential formulations was written
628 by Voter.\cite{voter}
629
630 The {\sc eam} potential has the form:
631 \begin{eqnarray}
632 V & = & \sum_{i} F_{i}\left[\rho_{i}\right] + \sum_{i} \sum_{j \neq i}
633 \phi_{ij}({\bf r}_{ij}) \\
634 \rho_{i} & = & \sum_{j \neq i} f_{j}({\bf r}_{ij})
635 \end{eqnarray}
636 where $F_{i} $ is the embedding function that equates the energy required to embed a
637 positively-charged core ion $i$ into a linear superposition of
638 spherically averaged atomic electron densities given by
639 $\rho_{i}$. $\phi_{ij}$ is a primarily repulsive pairwise interaction
640 between atoms $i$ and $j$. In the original formulation of
641 {\sc eam}\cite{Daw84}, $\phi_{ij}$ was an entirely repulsive term, however
642 in later refinements to EAM have shown that non-uniqueness between $F$
643 and $\phi$ allow for more general forms for $\phi$.\cite{Daw89}
644 There is a cutoff distance, $r_{cut}$, which limits the
645 summations in the {\sc eam} equation to the few dozen atoms
646 surrounding atom $i$ for both the density $\rho$ and pairwise $\phi$
647 interactions. Foiles et al. fit EAM potentials for fcc metals Cu, Ag, Au, Ni, Pd, Pt and alloys of these metals\cite{FBD86}. These potential fits are in the DYNAMO 86 format and are included with {\sc oopse}.
648
649
650 \subsection{\label{oopseSec:pbc}Periodic Boundary Conditions}
651
652 \newcommand{\roundme}{\operatorname{round}}
653
654 \textit{Periodic boundary conditions} are widely used to simulate bulk properties with a relatively small number of particles. The
655 simulation box is replicated throughout space to form an infinite
656 lattice. During the simulation, when a particle moves in the primary
657 cell, its image in other cells move in exactly the same direction with
658 exactly the same orientation. Thus, as a particle leaves the primary
659 cell, one of its images will enter through the opposite face. If the
660 simulation box is large enough to avoid ``feeling'' the symmetries of
661 the periodic lattice, surface effects can be ignored. The available
662 periodic cells in OOPSE are cubic, orthorhombic and parallelepiped. We
663 use a $3 \times 3$ matrix, $\mathbf{H}$, to describe the shape and
664 size of the simulation box. $\mathbf{H}$ is defined:
665 \begin{equation}
666 \mathbf{H} = ( \mathbf{h}_x, \mathbf{h}_y, \mathbf{h}_z )
667 \end{equation}
668 Where $\mathbf{h}_j$ is the column vector of the $j$th axis of the
669 box. During the course of the simulation both the size and shape of
670 the box can be changed to allow volume fluctations when constraining
671 the pressure.
672
673 A real space vector, $\mathbf{r}$ can be transformed in to a box space
674 vector, $\mathbf{s}$, and back through the following transformations:
675 \begin{align}
676 \mathbf{s} &= \mathbf{H}^{-1} \mathbf{r} \\
677 \mathbf{r} &= \mathbf{H} \mathbf{s}
678 \end{align}
679 The vector $\mathbf{s}$ is now a vector expressed as the number of box
680 lengths in the $\mathbf{h}_x$, $\mathbf{h}_y$, and $\mathbf{h}_z$
681 directions. To find the minimum image of a vector $\mathbf{r}$, we
682 first convert it to its corresponding vector in box space, and then,
683 cast each element to lie on the in the range $[-0.5,0.5]$:
684 \begin{equation}
685 s_{i}^{\prime}=s_{i}-\roundme(s_{i})
686 \end{equation}
687 Where $s_i$ is the $i$th element of $\mathbf{s}$, and
688 $\roundme(s_i)$is given by
689 \begin{equation}
690 \roundme(x) =
691 \begin{cases}
692 \lfloor x+0.5 \rfloor & \text{if $x \ge 0$} \\
693 \lceil x-0.5 \rceil & \text{if $x < 0$ }
694 \end{cases}
695 \end{equation}
696 Here $\lfloor x \rfloor$ is the floor operator, and gives the largest
697 integer value that is not greater than $x$, and $\lceil x \rceil$ is
698 the ceiling operator, and gives the smallest integer that is not less
699 than $x$. For example, $\roundme(3.6)=4$, $\roundme(3.1)=3$,
700 $\roundme(-3.6)=-4$, $\roundme(-3.1)=-3$.
701
702 Finally, we obtain the minimum image coordinates $\mathbf{r}^{\prime}$ by
703 transforming back to real space,
704 \begin{equation}
705 \mathbf{r}^{\prime}=\mathbf{H}^{-1}\mathbf{s}^{\prime}%
706 \end{equation}
707 In this way, particles are allowed to diffuse freely in $\mathbf{r}$,
708 but their minimum images, $\mathbf{r}^{\prime}$ are used to compute
709 the interatomic forces.
710
711
712 \section{\label{oopseSec:IOfiles}Input and Output Files}
713
714 \subsection{{\sc bass} and Model Files}
715
716 Every {\sc oopse} simulation begins with a {\sc bass} file. {\sc bass}
717 (\underline{B}izarre \underline{A}tom \underline{S}imulation
718 \underline{S}yntax) is a script syntax that is parsed by {\sc oopse} at
719 runtime. The {\sc bass} file allows for the user to completely describe the
720 system they are to simulate, as well as tailor {\sc oopse}'s behavior during
721 the simulation. {\sc bass} files are denoted with the extension
722 \texttt{.bass}, an example file is shown in
723 Fig.~\ref{fig:bassExample}.
724
725 \begin{figure}
726 \centering
727 \framebox[\linewidth]{\rule{0cm}{0.75\linewidth}I'm a {\sc bass} file!}
728 \caption{Here is an example \texttt{.bass} file}
729 \label{fig:bassExample}
730 \end{figure}
731
732 Within the \texttt{.bass} file it is necessary to provide a complete
733 description of the molecule before it is actually placed in the
734 simulation. The {\sc bass} syntax was originally developed with this goal in
735 mind, and allows for the specification of all the atoms in a molecular
736 prototype, as well as any bonds, bends, or torsions. These
737 descriptions can become lengthy for complex molecules, and it would be
738 inconvenient to duplicate the simulation at the beginning of each {\sc bass}
739 script. Addressing this issue {\sc bass} allows for the inclusion of model
740 files at the top of a \texttt{.bass} file. These model files, denoted
741 with the \texttt{.mdl} extension, allow the user to describe a
742 molecular prototype once, then simply include it into each simulation
743 containing that molecule.
744
745 \subsection{\label{oopseSec:coordFiles}Coordinate Files}
746
747 The standard format for storage of a systems coordinates is a modified
748 xyz-file syntax, the exact details of which can be seen in
749 App.~\ref{appCoordFormat}. As all bonding and molecular information is
750 stored in the \texttt{.bass} and \texttt{.mdl} files, the coordinate
751 files are simply the complete set of coordinates for each atom at a
752 given simulation time.
753
754 There are three major files used by {\sc oopse} written in the coordinate
755 format, they are as follows: the initialization file, the simulation
756 trajectory file, and the final coordinates of the simulation. The
757 initialization file is necessary for {\sc oopse} to start the simulation
758 with the proper coordinates. It is typically denoted with the
759 extension \texttt{.init}. The trajectory file is created at the
760 beginning of the simulation, and is used to store snapshots of the
761 simulation at regular intervals. The first frame is a duplication of
762 the \texttt{.init} file, and each subsequent frame is appended to the
763 file at an interval specified in the \texttt{.bass} file. The
764 trajectory file is given the extension \texttt{.dump}. The final
765 coordinate file is the end of run or \texttt{.eor} file. The
766 \texttt{.eor} file stores the final configuration of the system for a
767 given simulation. The file is updated at the same time as the
768 \texttt{.dump} file. However, it only contains the most recent
769 frame. In this way, an \texttt{.eor} file may be used as the
770 initialization file to a second simulation in order to continue or
771 recover the previous simulation.
772
773 \subsection{\label{oopseSec:initCoords}Generation of Initial Coordinates}
774
775 As was stated in Sec.~\ref{oopseSec:coordFiles}, an initialization file
776 is needed to provide the starting coordinates for a simulation. The
777 {\sc oopse} package provides a program called \texttt{sysBuilder} to aid in
778 the creation of the \texttt{.init} file. \texttt{sysBuilder} is {\sc bass}
779 aware, and will recognize arguments and parameters in the
780 \texttt{.bass} file that would otherwise be ignored by the
781 simulation. The program itself is under continual development, and is
782 offered here as a helper tool only.
783
784 \subsection{The Statistics File}
785
786 The last output file generated by {\sc oopse} is the statistics file. This
787 file records such statistical quantities as the instantaneous
788 temperature, volume, pressure, etc. It is written out with the
789 frequency specified in the \texttt{.bass} file. The file allows the
790 user to observe the system variables as a function of simulation time
791 while the simulation is in progress. One useful function the
792 statistics file serves is to monitor the conserved quantity of a given
793 simulation ensemble, this allows the user to observe the stability of
794 the integrator. The statistics file is denoted with the \texttt{.stat}
795 file extension.
796
797 \section{\label{oopseSec:mechanics}Mechanics}
798
799 \subsection{\label{oopseSec:integrate}Integrating the Equations of Motion: the Symplectic Step Integrator}
800
801 Integration of the equations of motion was carried out using the
802 symplectic splitting method proposed by Dullweber \emph{et
803 al.}.\cite{Dullweber1997} The reason for this integrator selection
804 deals with poor energy conservation of rigid body systems using
805 quaternions. While quaternions work well for orientational motion in
806 alternate ensembles, the microcanonical ensemble has a constant energy
807 requirement that is quite sensitive to errors in the equations of
808 motion. The original implementation of this code utilized quaternions
809 for rotational motion propagation; however, a detailed investigation
810 showed that they resulted in a steady drift in the total energy,
811 something that has been observed by others.\cite{Laird97}
812
813 The key difference in the integration method proposed by Dullweber
814 \emph{et al.} is that the entire rotation matrix is propagated from
815 one time step to the next. In the past, this would not have been as
816 feasible a option, being that the rotation matrix for a single body is
817 nine elements long as opposed to 3 or 4 elements for Euler angles and
818 quaternions respectively. System memory has become much less of an
819 issue in recent times, and this has resulted in substantial benefits
820 in energy conservation. There is still the issue of 5 or 6 additional
821 elements for describing the orientation of each particle, which will
822 increase dump files substantially. Simply translating the rotation
823 matrix into its component Euler angles or quaternions for storage
824 purposes relieves this burden.
825
826 The symplectic splitting method allows for Verlet style integration of
827 both linear and angular motion of rigid bodies. In the integration
828 method, the orientational propagation involves a sequence of matrix
829 evaluations to update the rotation matrix.\cite{Dullweber1997} These
830 matrix rotations end up being more costly computationally than the
831 simpler arithmetic quaternion propagation. With the same time step, a
832 1000 SSD particle simulation shows an average 7\% increase in
833 computation time using the symplectic step method in place of
834 quaternions. This cost is more than justified when comparing the
835 energy conservation of the two methods as illustrated in figure
836 \ref{timestep}.
837
838 \begin{figure}
839 \centering
840 \includegraphics[width=\linewidth]{timeStep.eps}
841 \caption{Energy conservation using quaternion based integration versus
842 the symplectic step method proposed by Dullweber \emph{et al.} with
843 increasing time step. For each time step, the dotted line is total
844 energy using the symplectic step integrator, and the solid line comes
845 from the quaternion integrator. The larger time step plots are shifted
846 up from the true energy baseline for clarity.}
847 \label{timestep}
848 \end{figure}
849
850 In figure \ref{timestep}, the resulting energy drift at various time
851 steps for both the symplectic step and quaternion integration schemes
852 is compared. All of the 1000 SSD particle simulations started with the
853 same configuration, and the only difference was the method for
854 handling rotational motion. At time steps of 0.1 and 0.5 fs, both
855 methods for propagating particle rotation conserve energy fairly well,
856 with the quaternion method showing a slight energy drift over time in
857 the 0.5 fs time step simulation. At time steps of 1 and 2 fs, the
858 energy conservation benefits of the symplectic step method are clearly
859 demonstrated. Thus, while maintaining the same degree of energy
860 conservation, one can take considerably longer time steps, leading to
861 an overall reduction in computation time.
862
863 Energy drift in these SSD particle simulations was unnoticeable for
864 time steps up to three femtoseconds. A slight energy drift on the
865 order of 0.012 kcal/mol per nanosecond was observed at a time step of
866 four femtoseconds, and as expected, this drift increases dramatically
867 with increasing time step. To insure accuracy in the constant energy
868 simulations, time steps were set at 2 fs and kept at this value for
869 constant pressure simulations as well.
870
871
872 \subsection{\label{sec:extended}Extended Systems for other Ensembles}
873
874
875 {\sc oopse} implements a
876
877
878 \subsubsection{\label{oopseSec:noseHooverThermo}Nose-Hoover Thermostatting}
879
880 To mimic the effects of being in a constant temperature ({\sc nvt})
881 ensemble, {\sc oopse} uses the Nose-Hoover extended system
882 approach.\cite{Hoover85} In this method, the equations of motion for
883 the particle positions and velocities are
884 \begin{eqnarray}
885 \dot{{\bf r}} & = & {\bf v} \\
886 \dot{{\bf v}} & = & \frac{{\bf f}}{m} - \chi {\bf v}
887 \label{eq:nosehoovereom}
888 \end{eqnarray}
889
890 $\chi$ is an ``extra'' variable included in the extended system, and
891 it is propagated using the first order equation of motion
892 \begin{equation}
893 \dot{\chi} = \frac{1}{\tau_{T}} \left( \frac{T}{T_{target}} - 1 \right)
894 \label{eq:nosehooverext}
895 \end{equation}
896 where $T_{target}$ is the target temperature for the simulation, and
897 $\tau_{T}$ is a time constant for the thermostat.
898
899 To select the Nose-Hoover {\sc nvt} ensemble, the {\tt ensemble = NVT;}
900 command would be used in the simulation's {\sc bass} file. There is
901 some subtlety in choosing values for $\tau_{T}$, and it is usually set
902 to values of a few ps. Within a {\sc bass} file, $\tau_{T}$ could be
903 set to 1 ps using the {\tt tauThermostat = 1000; } command.
904
905
906 \subsection{\label{oopseSec:zcons}Z-Constraint Method}
907
908 Based on fluctuation-dissipation theorem,\bigskip\ force auto-correlation
909 method was developed to investigate the dynamics of ions inside the ion
910 channels.\cite{Roux91} Time-dependent friction coefficient can be calculated
911 from the deviation of the instantaneous force from its mean force.
912
913 %
914
915 \begin{equation}
916 \xi(z,t)=\langle\delta F(z,t)\delta F(z,0)\rangle/k_{B}T
917 \end{equation}
918 where%
919 \begin{equation}
920 \delta F(z,t)=F(z,t)-\langle F(z,t)\rangle
921 \end{equation}
922
923
924 If the time-dependent friction decay rapidly, static friction coefficient can
925 be approximated by%
926
927 \begin{equation}
928 \xi^{static}(z)=\int_{0}^{\infty}\langle\delta F(z,t)\delta F(z,0)\rangle dt
929 \end{equation}
930
931
932 Hence, diffusion constant can be estimated by
933 \begin{equation}
934 D(z)=\frac{k_{B}T}{\xi^{static}(z)}=\frac{(k_{B}T)^{2}}{\int_{0}^{\infty
935 }\langle\delta F(z,t)\delta F(z,0)\rangle dt}%
936 \end{equation}
937
938
939 \bigskip Z-Constraint method, which fixed the z coordinates of the molecules
940 with respect to the center of the mass of the system, was proposed to obtain
941 the forces required in force auto-correlation method.\cite{Marrink94} However,
942 simply resetting the coordinate will move the center of the mass of the whole
943 system. To avoid this problem, a new method was used at {\sc oopse}. Instead of
944 resetting the coordinate, we reset the forces of z-constraint molecules as
945 well as subtract the total constraint forces from the rest of the system after
946 force calculation at each time step.
947 \begin{align}
948 F_{\alpha i}&=0\\
949 V_{\alpha i}&=V_{\alpha i}-\frac{\sum\limits_{i}M_{_{\alpha i}}V_{\alpha i}}{\sum\limits_{i}M_{_{\alpha i}}}\\
950 F_{\alpha i}&=F_{\alpha i}-\frac{M_{_{\alpha i}}}{\sum\limits_{\alpha}\sum\limits_{i}M_{_{\alpha i}}}\sum\limits_{\beta}F_{\beta}\\
951 V_{\alpha i}&=V_{\alpha i}-\frac{\sum\limits_{\alpha}\sum\limits_{i}M_{_{\alpha i}}V_{\alpha i}}{\sum\limits_{\alpha}\sum\limits_{i}M_{_{\alpha i}}}
952 \end{align}
953
954 At the very beginning of the simulation, the molecules may not be at its
955 constraint position. To move the z-constraint molecule to the specified
956 position, a simple harmonic potential is used%
957
958 \begin{equation}
959 U(t)=\frac{1}{2}k_{Harmonic}(z(t)-z_{cons})^{2}%
960 \end{equation}
961 where $k_{Harmonic}$\bigskip\ is the harmonic force constant, $z(t)$ is
962 current z coordinate of the center of mass of the z-constraint molecule, and
963 $z_{cons}$ is the restraint position. Therefore, the harmonic force operated
964 on the z-constraint molecule at time $t$ can be calculated by%
965 \begin{equation}
966 F_{z_{Harmonic}}(t)=-\frac{\partial U(t)}{\partial z(t)}=-k_{Harmonic}%
967 (z(t)-z_{cons})
968 \end{equation}
969 Worthy of mention, other kinds of potential functions can also be used to
970 drive the z-constraint molecule.
971
972 \section{\label{oopseSec:props}Trajectory Analysis}
973
974 \subsection{\label{oopseSec:staticProps}Static Property Analysis}
975
976 The static properties of the trajectories are analyzed with the
977 program \texttt{staticProps}. The code is capable of calculating the following
978 pair correlations between species A and B:
979 \begin{itemize}
980 \item $g_{\text{AB}}(r)$: Eq.~\ref{eq:gofr}
981 \item $g_{\text{AB}}(r, \cos \theta)$: Eq.~\ref{eq:gofrCosTheta}
982 \item $g_{\text{AB}}(r, \cos \omega)$: Eq.~\ref{eq:gofrCosOmega}
983 \item $g_{\text{AB}}(x, y, z)$: Eq.~\ref{eq:gofrXYZ}
984 \item $\langle \cos \omega \rangle_{\text{AB}}(r)$:
985 Eq.~\ref{eq:cosOmegaOfR}
986 \end{itemize}
987
988 The first pair correlation, $g_{\text{AB}}(r)$, is defined as follows:
989 \begin{equation}
990 g_{\text{AB}}(r) = \frac{V}{N_{\text{A}}N_{\text{B}}}\langle %%
991 \sum_{i \in \text{A}} \sum_{j \in \text{B}} %%
992 \delta( r - |\mathbf{r}_{ij}|) \rangle \label{eq:gofr}
993 \end{equation}
994 Where $\mathbf{r}_{ij}$ is the vector
995 \begin{equation*}
996 \mathbf{r}_{ij} = \mathbf{r}_j - \mathbf{r}_i \notag
997 \end{equation*}
998 and $\frac{V}{N_{\text{A}}N_{\text{B}}}$ normalizes the average over
999 the expected pair density at a given $r$.
1000
1001 The next two pair correlations, $g_{\text{AB}}(r, \cos \theta)$ and
1002 $g_{\text{AB}}(r, \cos \omega)$, are similar in that they are both two
1003 dimensional histograms. Both use $r$ for the primary axis then a
1004 $\cos$ for the secondary axis ($\cos \theta$ for
1005 Eq.~\ref{eq:gofrCosTheta} and $\cos \omega$ for
1006 Eq.~\ref{eq:gofrCosOmega}). This allows for the investigator to
1007 correlate alignment on directional entities. $g_{\text{AB}}(r, \cos
1008 \theta)$ is defined as follows:
1009 \begin{equation}
1010 g_{\text{AB}}(r, \cos \theta) = \frac{V}{N_{\text{A}}N_{\text{B}}}\langle
1011 \sum_{i \in \text{A}} \sum_{j \in \text{B}}
1012 \delta( \cos \theta - \cos \theta_{ij})
1013 \delta( r - |\mathbf{r}_{ij}|) \rangle
1014 \label{eq:gofrCosTheta}
1015 \end{equation}
1016 Where
1017 \begin{equation*}
1018 \cos \theta_{ij} = \mathbf{\hat{i}} \cdot \mathbf{\hat{r}}_{ij}
1019 \end{equation*}
1020 Here $\mathbf{\hat{i}}$ is the unit directional vector of species $i$
1021 and $\mathbf{\hat{r}}_{ij}$ is the unit vector associated with vector
1022 $\mathbf{r}_{ij}$.
1023
1024 The second two dimensional histogram is of the form:
1025 \begin{equation}
1026 g_{\text{AB}}(r, \cos \omega) =
1027 \frac{V}{N_{\text{A}}N_{\text{B}}}\langle
1028 \sum_{i \in \text{A}} \sum_{j \in \text{B}}
1029 \delta( \cos \omega - \cos \omega_{ij})
1030 \delta( r - |\mathbf{r}_{ij}|) \rangle \label{eq:gofrCosOmega}
1031 \end{equation}
1032 Here
1033 \begin{equation*}
1034 \cos \omega_{ij} = \mathbf{\hat{i}} \cdot \mathbf{\hat{j}}
1035 \end{equation*}
1036 Again, $\mathbf{\hat{i}}$ and $\mathbf{\hat{j}}$ are the unit
1037 directional vectors of species $i$ and $j$.
1038
1039 The static analysis code is also cable of calculating a three
1040 dimensional pair correlation of the form:
1041 \begin{equation}\label{eq:gofrXYZ}
1042 g_{\text{AB}}(x, y, z) =
1043 \frac{V}{N_{\text{A}}N_{\text{B}}}\langle
1044 \sum_{i \in \text{A}} \sum_{j \in \text{B}}
1045 \delta( x - x_{ij})
1046 \delta( y - y_{ij})
1047 \delta( z - z_{ij}) \rangle
1048 \end{equation}
1049 Where $x_{ij}$, $y_{ij}$, and $z_{ij}$ are the $x$, $y$, and $z$
1050 components respectively of vector $\mathbf{r}_{ij}$.
1051
1052 The final pair correlation is similar to
1053 Eq.~\ref{eq:gofrCosOmega}. $\langle \cos \omega
1054 \rangle_{\text{AB}}(r)$ is calculated in the following way:
1055 \begin{equation}\label{eq:cosOmegaOfR}
1056 \langle \cos \omega \rangle_{\text{AB}}(r) =
1057 \langle \sum_{i \in \text{A}} \sum_{j \in \text{B}}
1058 (\cos \omega_{ij}) \delta( r - |\mathbf{r}_{ij}|) \rangle
1059 \end{equation}
1060 Here $\cos \omega_{ij}$ is defined in the same way as in
1061 Eq.~\ref{eq:gofrCosOmega}. This equation is a single dimensional pair
1062 correlation that gives the average correlation of two directional
1063 entities as a function of their distance from each other.
1064
1065 All static properties are calculated on a frame by frame basis. The
1066 trajectory is read a single frame at a time, and the appropriate
1067 calculations are done on each frame. Once one frame is finished, the
1068 next frame is read in, and a running average of the property being
1069 calculated is accumulated in each frame. The program allows for the
1070 user to specify more than one property be calculated in single run,
1071 preventing the need to read a file multiple times.
1072
1073 \subsection{\label{dynamicProps}Dynamic Property Analysis}
1074
1075 The dynamic properties of a trajectory are calculated with the program
1076 \texttt{dynamicProps}. The program will calculate the following properties:
1077 \begin{gather}
1078 \langle | \mathbf{r}(t) - \mathbf{r}(0) |^2 \rangle \label{eq:rms}\\
1079 \langle \mathbf{v}(t) \cdot \mathbf{v}(0) \rangle \label{eq:velCorr} \\
1080 \langle \mathbf{j}(t) \cdot \mathbf{j}(0) \rangle \label{eq:angularVelCorr}
1081 \end{gather}
1082
1083 Eq.~\ref{eq:rms} is the root mean square displacement
1084 function. Eq.~\ref{eq:velCorr} and Eq.~\ref{eq:angularVelCorr} are the
1085 velocity and angular velocity correlation functions respectively. The
1086 latter is only applicable to directional species in the simulation.
1087
1088 The \texttt{dynamicProps} program handles he file in a manner different from
1089 \texttt{staticProps}. As the properties calculated by this program are time
1090 dependent, multiple frames must be read in simultaneously by the
1091 program. For small trajectories this is no problem, and the entire
1092 trajectory is read into memory. However, for long trajectories of
1093 large systems, the files can be quite large. In order to accommodate
1094 large files, \texttt{dynamicProps} adopts a scheme whereby two blocks of memory
1095 are allocated to read in several frames each.
1096
1097 In this two block scheme, the correlation functions are first
1098 calculated within each memory block, then the cross correlations
1099 between the frames contained within the two blocks are
1100 calculated. Once completed, the memory blocks are incremented, and the
1101 process is repeated. A diagram illustrating the process is shown in
1102 Fig.~\ref{oopseFig:dynamicPropsMemory}. As was the case with
1103 \texttt{staticProps}, multiple properties may be calculated in a
1104 single run to avoid multiple reads on the same file.
1105
1106
1107
1108 \section{\label{oopseSec:design}Program Design}
1109
1110 \subsection{\label{sec:architecture} {\sc oopse} Architecture}
1111
1112 The core of OOPSE is divided into two main object libraries:
1113 \texttt{libBASS} and \texttt{libmdtools}. \texttt{libBASS} is the
1114 library developed around the parsing engine and \texttt{libmdtools}
1115 is the software library developed around the simulation engine. These
1116 two libraries are designed to encompass all the basic functions and
1117 tools that {\sc oopse} provides. Utility programs, such as the
1118 property analyzers, need only link against the software libraries to
1119 gain access to parsing, force evaluation, and input / output
1120 routines.
1121
1122 Contained in \texttt{libBASS} are all the routines associated with
1123 reading and parsing the \texttt{.bass} input files. Given a
1124 \texttt{.bass} file, \texttt{libBASS} will open it and any associated
1125 \texttt{.mdl} files; then create structures in memory that are
1126 templates of all the molecules specified in the input files. In
1127 addition, any simulation parameters set in the \texttt{.bass} file
1128 will be placed in a structure for later query by the controlling
1129 program.
1130
1131 Located in \texttt{libmdtools} are all other routines necessary to a
1132 Molecular Dynamics simulation. The library uses the main data
1133 structures returned by \texttt{libBASS} to initialize the various
1134 parts of the simulation: the atom structures and positions, the force
1135 field, the integrator, \emph{et cetera}. After initialization, the
1136 library can be used to perform a variety of tasks: integrate a
1137 Molecular Dynamics trajectory, query phase space information from a
1138 specific frame of a completed trajectory, or even recalculate force or
1139 energetic information about specific frames from a completed
1140 trajectory.
1141
1142 With these core libraries in place, several programs have been
1143 developed to utilize the routines provided by \texttt{libBASS} and
1144 \texttt{libmdtools}. The main program of the package is \texttt{oopse}
1145 and the corresponding parallel version \texttt{oopse\_MPI}. These two
1146 programs will take the \texttt{.bass} file, and create then integrate
1147 the simulation specified in the script. The two analysis programs
1148 \texttt{staticProps} and \texttt{dynamicProps} utilize the core
1149 libraries to initialize and read in trajectories from previously
1150 completed simulations, in addition to the ability to use functionality
1151 from \texttt{libmdtools} to recalculate forces and energies at key
1152 frames in the trajectories. Lastly, the family of system building
1153 programs (Sec.~\ref{oopseSec:initCoords}) also use the libraries to
1154 store and output the system configurations they create.
1155
1156 \subsection{\label{oopseSec:parallelization} Parallelization of {\sc oopse}}
1157
1158 Although processor power is continually growing month by month, it is
1159 still unreasonable to simulate systems of more then a 1000 atoms on a
1160 single processor. To facilitate study of larger system sizes or
1161 smaller systems on long time scales in a reasonable period of time,
1162 parallel methods were developed allowing multiple CPU's to share the
1163 simulation workload. Three general categories of parallel
1164 decomposition method's have been developed including atomic, spatial
1165 and force decomposition methods.
1166
1167 Algorithmically simplest of the three method's is atomic decomposition
1168 where N particles in a simulation are split among P processors for the
1169 duration of the simulation. Computational cost scales as an optimal
1170 $O(N/P)$ for atomic decomposition. Unfortunately all processors must
1171 communicate positions and forces with all other processors leading
1172 communication to scale as an unfavorable $O(N)$ independent of the
1173 number of processors. This communication bottleneck led to the
1174 development of spatial and force decomposition methods in which
1175 communication among processors scales much more favorably. Spatial or
1176 domain decomposition divides the physical spatial domain into 3D boxes
1177 in which each processor is responsible for calculation of forces and
1178 positions of particles located in its box. Particles are reassigned to
1179 different processors as they move through simulation space. To
1180 calculate forces on a given particle, a processor must know the
1181 positions of particles within some cutoff radius located on nearby
1182 processors instead of the positions of particles on all
1183 processors. Both communication between processors and computation
1184 scale as $O(N/P)$ in the spatial method. However, spatial
1185 decomposition adds algorithmic complexity to the simulation code and
1186 is not very efficient for small N since the overall communication
1187 scales as the surface to volume ratio $(N/P)^{2/3}$ in three
1188 dimensions.
1189
1190 Force decomposition assigns particles to processors based on a block
1191 decomposition of the force matrix. Processors are split into a
1192 optimally square grid forming row and column processor groups. Forces
1193 are calculated on particles in a given row by particles located in
1194 that processors column assignment. Force decomposition is less complex
1195 to implement then the spatial method but still scales computationally
1196 as $O(N/P)$ and scales as $(N/\sqrt{p})$ in communication
1197 cost. Plimpton also found that force decompositions scales more
1198 favorably then spatial decomposition up to 10,000 atoms and favorably
1199 competes with spatial methods for up to 100,000 atoms.
1200
1201 \subsection{\label{oopseSec:memAlloc}Memory Issues in Trajectory Analysis}
1202
1203 For large simulations, the trajectory files can sometimes reach sizes
1204 in excess of several gigabytes. In order to effectively analyze that
1205 amount of data+, two memory management schemes have been devised for
1206 \texttt{staticProps} and for \texttt{dynamicProps}. The first scheme,
1207 developed for \texttt{staticProps}, is the simplest. As each frame's
1208 statistics are calculated independent of each other, memory is
1209 allocated for each frame, then freed once correlation calculations are
1210 complete for the snapshot. To prevent multiple passes through a
1211 potentially large file, \texttt{staticProps} is capable of calculating
1212 all requested correlations per frame with only a single pair loop in
1213 each frame and a single read through of the file.
1214
1215 The second, more advanced memory scheme, is used by
1216 \texttt{dynamicProps}. Here, the program must have multiple frames in
1217 memory to calculate time dependent correlations. In order to prevent a
1218 situation where the program runs out of memory due to large
1219 trajectories, the user is able to specify that the trajectory be read
1220 in blocks. The number of frames in each block is specified by the
1221 user, and upon reading a block of the trajectory,
1222 \texttt{dynamicProps} will calculate all of the time correlation frame
1223 pairs within the block. After in block correlations are complete, a
1224 second block of the trajectory is read, and the cross correlations are
1225 calculated between the two blocks. this second block is then freed and
1226 then incremented and the process repeated until the end of the
1227 trajectory. Once the end is reached, the first block is freed then
1228 incremented, and the again the internal time correlations are
1229 calculated. The algorithm with the second block is then repeated with
1230 the new origin block, until all frame pairs have been correlated in
1231 time. This process is illustrated in
1232 Fig.~\ref{oopseFig:dynamicPropsMemory}.
1233
1234 \begin{figure}
1235 \centering
1236 \includegraphics[width=\linewidth]{dynamicPropsMem.eps}
1237 \caption[A representation of the block correlations in \texttt{dynamicProps}]{This diagram illustrates the memory management used by \texttt{dynamicProps}, which follows the scheme: $\sum^{N_{\text{memory blocks}}}_{i=1}[ \operatorname{self}(i) + \sum^{N_{\text{memory blocks}}}_{j>i} \operatorname{cross}(i,j)]$. The shaded region represents the self correlation of the memory block, and the open blocks are read one at a time and the cross correlations between blocks are calculated.}
1238 \label{oopseFig:dynamicPropsMemory}
1239 \end{figure}
1240
1241 \subsection{\label{openSource}Open Source and Distribution License}
1242
1243 \section{\label{oopseSec:conclusion}Conclusion}
1244
1245 We have presented the design and implementation of our open source
1246 simulation package {\sc oopse}. The package offers novel
1247 capabilities to the field of Molecular Dynamics simulation packages in
1248 the form of dipolar force fields, and symplectic integration of rigid
1249 body dynamics. It is capable of scaling across multiple processors
1250 through the use of MPI. It also implements several integration
1251 ensembles allowing the end user control over temperature and
1252 pressure. In addition, it is capable of integrating constrained
1253 dynamics through both the {\sc rattle} algorithm and the z-constraint
1254 method.
1255
1256 These features are all brought together in a single open-source
1257 development package. This allows researchers to not only benefit from
1258 {\sc oopse}, but also contribute to {\sc oopse}'s development as
1259 well.Documentation and source code for {\sc oopse} can be downloaded
1260 from \texttt{http://www.openscience.org/oopse/}.
1261