Miscellaneous and Debugging Opcodes

The opcodes in this section are linked below:

DEBG — DEBuGging

DUMP — DUMP connection table

GEOM — obtain GEOMetric information about the molecule

TIME — report cpu TIME (user+system)

JWRT — Journal WRiTe

JRED — Journal REaD

NPRC — Number of PRoCessors (distributed MacroModel calculations)

SPAT — SPecial Atom Treatment

`DEBG` — DEBuGging

Turn on program debug switch (numbered 1-1100). Use as many of these commands as necessary to turn on as many of the switches as you want.

arg1-4	Debug switches
	Values currently in use are listed below. “GV” stands for “Greater Verbosity.” Flags described with this abbreviation affect only the level of reporting; other flags affect program action.
1	GV, all commands
2	Don’t line-buffer `.log` file
3	Always write full (never compressed) output files
5	GV, dynamic allocation
6	Always write old-style (mmio) files
7	Always write new-style (m2io) files
8	GV, correspondence between mainCT and sbstrCT numbering
9	GV, handling of mmio/m2io files
10	GV, RWND command
11	Get formal charge from `.fld` instead of `atom.typ`
12	Don’t set qq product to zero, VDWB nonbondeds
13	GV, LIGB, and VDWB bonds
14	GV, SUBS, FXAT, FXDI, FXBA, FXTA command handlers
15	GV, reading charges from ffld substructures
16	GV, VDW offset from ffld
17	Suppress elimination of constrained-atom mutual interactions
18	GV, charge delocalization
19	GV, number of matches, each ffld substructure in mol
20	GV, name of each ffld substr matched in mol, and atom #s
21	GV, more details on how each ffld substr matched mol
22	GV, write problematic structure to output structure file if MINI fails
23	GV, write MINI structure to output structure file at each MINI display interval. Incompatible with DRIV.
25	Don’t refuse to minimize extremely strained structures
27	GV, TNCG, FMNR MINI methods
28	Suppress elimination of stretching and bending interactions in which some of the atoms are fixed.
30	GV, MINI line search
31	GV, MINI gradient to log file, each iteration
32	Don’t use constant derivatives in line searching
33	Don’t separate close pairs during interaction generation
34	GV, tests for distorted sp3 atom and for chirality
35	Do write output `.sbc` file
36	GV, do not limit the number of messages from close separation problems
37	Note that if separation of close pairs fails in a conformational search reject the structure and continue the search
38	Skip tests for distorted sp³ atoms.
40	GV, early in force-field (eqn. reading, etc.)
41	GV, BMFF IPC layer (client and server side)
42	GV, BMFF server
43	GV, BMFF-related stuff within MacroModel
44	GV, applying mmlewis and obtaining OPLS_2005 parameters
45	Do not delete dummy atoms for OPLS4. Dummy atoms are correctly handled in OPLS4.
46	Don’t use fast vdw and hbd BMFF processing
49	Don’t skip over structures that do not have all force field and solvation parameters. Stop when such a structure is encountered.
50	GV, parsing of `atom.typ` file
51	LV, parameter assignment
55	Don’t compare the connectivity of the input structure with the previous structure, but treat it as chemically different from the previous structure.
56	Don’t eliminate torsional interactions with V’s of 0
57	Always consider input structures to be distinct CT’s, not conformations. This implies full interaction generation for each structure read in.
58	GV, when which classes of params are updated (Bonded, VDW, Geom-dependent, solvation)
59	GV, geom-dependent params & updating
60	GV, `.fld` ALT and SEL processing
61	GV, stretch interactions
62	GV, bend interactions
63	GV, torsional interactions
64	GV, nonbonded interactions
65	GV, bend-bend interactions
68	GV, `.fld` atom-type equivalencing
69	GV, `.fld` line-numbering & assembly of interactions; to `.mmo`
70	GV, fixed atom interactions
71	GV, parameters that change during FEP
72	GV, atoms whose surface areas change during FEP
73	Don’t eliminate str/bend involving dummy atoms or zero-order bonds in	FEP accumulation
74	Exclude all str/bend terms from FEP accumulation
78	Calculate normal modes only for SUBS atoms in LMOD and LLMOD calculations
79	ATEQ, also consider noncyclic permutations of the `ATEQ` atoms
81	GV, ATEQ, during csearch or mult MINI test for uniqueness
82	GV, Comparing structures for uniqueness
83	GV, MOLS and TORS moving-atom sets
85	Use `DRVPOL_OLD` instead of `DRVPOL`; `DRVPOL` is faster for 1st derivs but has bad 2nd derivs. We use `DRVPOL_OLD` when we need 2nd derivs or when `DEBG 85` is specified
86	Use current rather than ideal str and bend distances in computing constant part of analytical GB radius
87	GV, atom-wise solvation data to `.mmo` even if E(atom)=0
88	GV, read `.slv` solvation file and match its atoms to mol
89	Don’t unite all-atom sp3 CHn groups for solvation
90	Include explicit hydrogens in GB-radius calculation
91	Use old (Hasel et al.) method for analytical surfaces
92	GV, low-mode search (LMCS or LMC2)
93	Turn off surface-area 3-body function
94	GV, saddle point search. Also, write intermediate structures to output structure file.
95	Use all-atom representation in surface calcs (assuming molecule actually has H’s and/or lone pairs)
96	Ignore long range shells in numerical Born rad. calc
97	GV, solvation energy calc
98	GV, parameters for overlap array, solvation calc
99	GV, analytical and numerical GB radii and Gpol values for fixed atoms
100	Set negative areas to zero in LCPO surface calculation
101	GV, MD or MC monitor of PE avg, sd, skew and kurtosis
102	GV, timings, MD and MINI
103	Turn off adaptive mechanism for number of Monte Carlo degrees of freedom during MCSD, MCLO, IMPS
104	After MD “Abort,” continue with remainder of `.com` file
105	Do not do LCPO Buried Atom Elimination (BAE)
106	GV, SHAKE
107	Do not do neighbor-list reduction (NLR) on LCPO overlap array
108	GV, write LCPO ELST `3` results to filename`.lcp` as well as `.mmo`. If `DEBG 91` is in effect, still write the file, which includes the numerical atomic surfaces, but with LCPO sums set to 0.
111	GV, general dynamics & MC
112	GV, write every 200th non-matching IMPS structure to output structure file
121	GV, BDCO pairlist generation
126	GV, BDCO total charge product
127	GV, BDCO pairwise charge product assignment
128	GV, more BDCO pairwise charge product assignment
129	GV, force field substructure explicit partial charge processing
130	GV, CHGF input structure file explicit partial charge to delocalized formal charge + bond charge increment decomposition
131	GV, details of lone-pair requirements for current ffld, mol
141	GV, structures accepted, etc., MC csearch
150	During dihedral driving start each incremental minimization from the initial structure as read from the filename`.mae` file
152	GV, interaction array changes from nonbonded update, MINI
169	GV, torsional angles stored by torsional memory
178	Use MacroModel-supplied random-number function
179	GV, print seed to log file each time random-number generator called
180	GV, movements used to generate new Monte Carlo structures
181	GV, internal coords used in tors MC and MCSD simulations
182	GV, internal coords, final limits after tors MC run
183	GV, internal coord changes at each tors MC acceptance
184	In MC, calculate total E, not just MC-variable E components
185	Allow JBW in substructures—for internal use only
186	Don’t reorient each output structure to best superimpose on input
187	GV, For IMPS, write each input structure to output, after zmat superposition of the first three atoms onto those of the first structure. Requires user specification of `DEBG 186`.
188	Skip problematic ring closure rotations. This may result in chirality switches if CHIG commands are not applied to the two central atoms in the ring closure.
191	Don’t record OPLS-AA formal charges or bond orders in the output structure file. Use the ones from the input structure file.
200	GV, CGEN searches
202	Override force field setting for ConfGen.
210	GV, mass-weighted force-constants, RRHO command
211	Allow RRHO, MTST on structures not necessarily minimized
222	GV, write starting structure for each MINI to output structure file
230	GV, COPY opcode
235	GV, ALGN opcode, general
236	GV, ALGN opcode, center of mass operations
237	GV, ALGN opcode, principal axis operations
251	GV, “torsional memory” during conformation search
252	GV, conformational search of ring structures
333	GV, network-distributed processing using the `NPRC` command
358	GV, geometrical pre-optimization of ring structures
360	Turn off use-directed csearch initialization from filename`.mae` file
370	Enable MOLS for molecules containing fixed atoms
400	GV, importance sampling
500	GV, source CT number tracking for Maestro project facility properties
510	Do not enforce properties dependencies (i.e., do not clear properties when the structure changes) In Embrace calculations it is turned on by default. Use arg4 of MBAE to turn it off.
511	GV, the process for recording properties
512	Do not add dependencies to MacroModel Properties that lack them.
513	Record energy components in the output structure file. This flag is on by default; use `DEBG` 514 to turn it off.
514	Do not record energy components in output structure file. Turns `DEBG` 513 off.
520	Basic reporting for Automatic setup (AUTO)
521	GV, reporting for Automatic setup (AUTO) and turn on `DEBG 520`
530	Record the structure for each molecule in both solvents for `LOGP` calculations in the output structure file
531	GV, for LOGP calculations
550	GV, MBAE
555	GV, LOOP
570	Don’t turn on `MMSYM_IN_PLACE` when a fixed or frozen atom is added. The use of `DEBG 570` is not recommended because `mmsym` comparisons which are not done in place translate and rotate the system without taking into account fixed or frozen atom positioning, resulting in distorted structures.
601	Turn off writing to output structure file unless explicit WRIT is used
602	If old-style titles are detected in a Maestro-formatted input structure file, do not update them.
720	GV, 1st and second derivs of Wilson angle OOP terms
725	Turn off constant long-range nonbonded derivs
726	Turn off constant long-range polarization solvation terms
727	Turn off constant long-range surface-area solvation terms
800	Don’t regenerate solvation overlap array after MC accept
820	Set all atomic weights to 12 for dynamics, inc. FEP
825	Don’t correct tiny denom. in analytical solvation derivs
830	Don’t correct Gpol,i' for nonbonded pairs left off the nonbonded pairlist
832	Don’t correct Gpol,i' for F-F nonbonded pairs which are farther apart than the nonbonded cutoff
	F = (a) fixed without flat bottom well or (b) frozen
835	GV, final Gpol,i' values for all atoms
836	GV, generalized Born CCF (Close Contact Function). Available only with single precision energies)
899	GV, FlexLM licensing general
900	GV, print RCS versions of source files used in compilation
901	GV, detailed tables of COMMON memory utilization
902	GV, table of atom info (regurgitation of `atom.typ` info)
903	GV, table of atomic masses by atom number
920	Low-mode search (LMCS or LMC2), write perturbed structures in output structure file
930	Do not check for `.stp`, `.upt`, `.slp` files (speeds response, especially on NFS-mounted filesystems, but loses the functionality of these checks). This behavior is on by default. To enable checking for `.stp`, `.upt`, `.slp` files, use `DEBG 931`.
931	Turns `DEBG 930` off, thus turning on checking for `.stp`, `.upt`, `.slp` files. Note that these files must be in the directory from which the MacroModel job is being run. It may be useful to use `DEBG 931` with the `-LOCAL` command-line option to MacroModel, which keeps job files in the current directory.
940	GV, for network-distributed MacroModel jobs using the `NPRC` command, save files written by slave processes. You must use the `-LOCAL` option to the `bmin` launch script when using this debug option.
950	GV, on IBM SP2, save file written by parallel threads
960	Turn off reporting of format problems in the `.com` file
961	Cause MacroModel to stop if a format problem is found in the `.com` file. This overrides `DEBG 960` and forces reporting of format problems.
975	Do not use atomic number information from the input file. Instead, obtain the atomic number information from the atom type.
999	GV, ARPK
1000	GV, give CPU timings in energetic routines. Now gives more detailed and informative output.
1001	GV, MINTA free-energy calculation.
1002	Turn off distance check on 1-3 atoms. Structures in which two atoms that are bonded to a third atom but not to each other are rejected by default if the two atoms are closer than 1 Å.

`DUMP` — DUMP connection table

List the connection table to the log file (used primarily for testing purposes).

`GEOM` — obtain GEOMetric information about the molecule

The GEOM command allows the user to obtain geometric information about the molecule from the MacroModel command file. Bond lengths, bond angles, and dihedral angles are written as properties to the log file.

arg1	Atom 1
	If arg1 is nonzero and arg2 is zero, then `GEOM` prints out the x, y, and z coordinates of the atom given by arg1.
arg2	Atom 2
	If arg1 and arg2 are nonzero and arg3 and arg5 are zero, `GEOM` prints the distance between the atoms specified by arg1 and arg2.
arg3	Atom 3
	If arg1, arg2, and arg3 are nonzero and arg4 is zero, `GEOM` prints the angle between the atoms specified in arg1, arg2, and arg3. If arg5 is nonzero, it is ignored.
arg4	Atom 4
	If arg1, arg2, arg3, and arg4 are nonzero, `GEOM` prints the dihedral angle between the atoms specified in arg1, arg2, arg3, and arg4. If arg5 is nonzero, it is ignored.
arg5	Spin-spin coupling constant
	If arg5 > 0, `GEOM` prints the spin-spin coupling constant J between the atoms specified in arg1 and arg2, subject to the following conditions: The atoms given by arg1 and arg2 are hydrogen atoms The hydrogen atoms are connected through two carbon atoms The values of arg3 and arg4 are zero.
	If either of the first two conditions is not met, a warning is printed and the calculation continues without calculating the coupling constant.

`TIME` — report cpu TIME (user+system)

The TIME command reports CPU time (user+system) since the previous invocation of TIME, or (on the first invocation) since program start. See also DEBG 1000 for more detailed and informative output on task timings.

`JWRT` — Journal WRiTe

The JRED and JWRT commands are provided in order to use MacroModel as a force-field server in connection with procedures being carried out by other processes. JWRT writes out a user-specifiable combination of coordinates, energy, gradient, and Hessian in binary form that the co-process can read. JRED reads in coordinates in a similar format that have been written out by the co-process. Thus, the co-process can provide coordinates to MacroModel, instruct it to compute the energetic terms, read back those energetic terms, manipulate the molecule, and complete the cycle. The way this facility is provided now, MacroModel has to be launched anew for each new calculation (for example, by a system() call from the client process); however, we may later enhance this facility to make it more fully interactive.

arg1	Write coordinates
	If arg1 is nonzero, write out the coordinates.
arg2	Write energy
	If arg2 is nonzero, write out the energy.
arg3	Write gradient
	If arg3 is nonzero, write out the gradient.
arg4	Write Hessian
	If arg4 is nonzero, write out the Hessian.

The filename used is filename.jwr, where filename is the prefix of the input filename. The file is opened with FORM=UNFORMATTED, but is sequential. All variables are written in double precision. Thus, if all four args are nonzero, and if there are N atoms in the part of the system being simulated (the whole system, unless SUBS is used with FXAT and all atoms are not covered by these commands), the .jwr file will contain 3N coordinates, in sequence x(1), y(1), z(1), x(2), ..., z(N), followed by one double precision value for the energy, followed by 3N gradient components, followed by (3N)² Hessian elements, in order x(1)x(1), y(1)x(1), z(1)x(1), x(2)x(1), ..., y(N)z(N), z(N)z(N).

`JRED` — Journal REaD

arg1	Read coordinates
	If arg1 is nonzero, read the coordinates.
arg2	Read energy
	If arg2 is nonzero, read the energy.
arg3	Read gradient
	If arg3 is nonzero, read the gradient.
arg4	Read Hessian
	If arg4 is nonzero, read the Hessian.
arg5	Print values read
	If arg5 is nonzero, whatever is read will be printed to the `.log` file. However, only the coordinates are used by subsequent computations in the program; the energy, gradient, and Hessian are simply read and printed, if requested.

JRED works as follows. It expects to read a file called filename.jrd, where filename is the stem of the input file name. For the purpose of debugging the JWRT command, args1-4 cause JRED to read the coordinates, the energy, the gradient, and the Hessian, respectively, if nonzero. If arg5 is nonzero, whatever is read will be printed to the .log file.

`NPRC` — Number of PRoCessors (distributed MacroModel calculations)

Distribute the MacroModel job over a number of different hosts. The name of the hosts to use are taken from the file schrodinger.hosts, which must be in the $SCHRODINGER directory or the user’s working directory. The NPRC command can be used with the following kinds of calculations:

Non-serial conformational searches
Serial conformational searches
MULT minimizations
Free-energy perturbation calculations (FEAV, FESA)
Embrace (MBAE) minimization and conformational search calculations

The types of non-serial conformational searches supported are: MCMM, LMCS, LMC2, mixed LMCS/MCMM, and mixed LMC2/MCMM. The types of serial searches supported are MCMM, LMCS, and mixed LMCS/MCMM. The AUTO command is required for serial MCMM and serial mixed LMCS/MCMM jobs, and is recommended for serial LMCS jobs. The procedure is split into a number of different searches and run on the hosts as they become available. In an MCMM procedure, a different SEED value is used on each host.

For MBAE energy difference calculations the initial receptor computations are conducted by the parent process prior to distributing the calculations for the ligands amongst the child processes. Because of the complexity of MBAE computations users are encouraged to consult the description of the MBAE opcode in this manual and the examples of MBAE calculations in Command File Examples for Embrace.

See Running Remote and Distributed MacroModel Calculations for a description of internally-distributed MacroModel calculations which use NPRC and see Preparing for Remote Job Submission (DEPRECATED) for information on setting computers and accounts to permit distributed MacroModel calculations.

If a requested host becomes unavailable during a run, its status is checked periodically. If the host later becomes available, the process starts using it again. Exception: hosts unavailable at startup time are removed from the list of allowed hosts and never re-used. This is because a host that was unavailable at the start of the job cannot have been checked for energetic consistency (see arg4 below), and using it without this test having been performed is potentially dangerous.

DEBG 333 gives greater verbosity; DEBG 940 saves, rather than discards, the output files of the intermediate jobs.

arg1	Number of hosts to check for availability
	arg1 specifies how many hosts in the `schrodinger.hosts` file (see Submitting MacroModel Jobs From the Command Line), will be checked for availability for a specific distributed run. If any hosts are not available, they are removed from the list of hosts to be used for the specified job. Each host that is checked but unavailable decreases the number of hosts over which the job will be distributed. Only hosts read and available will be used.
	For example, if the `schrodinger.hosts` file lists ten hosts, and arg1 is set to 5, then the first five hosts listed in the `scmchrodinger.hosts` file will be checked for availability. If two of the first five hosts are unavailable, the job will be distributed over three hosts.
arg2	Job size
	For non-serial searches arg2 specifies the number of steps from the search that should be carried out by each job. The total number of steps is the number given by arg1 of the search command (i.e. arg1 of MCMM, LMCS, or LMC2) and the total number of jobs is roughly given by arg1 of the search command divided by arg2 of `NPRC`.
	For serial calculations, such as minimizations of multiple input structures and serial searches this argument specifies the number of input structures given to each job for minimization or searching. In serial searches separate searches are conducted for each structure in the input structure file. Serial searches can be conducted for MCMM, LMCS, LMC2, mixed LMCS`/`MCMM and SPMC searches. The total number of jobs is given by the number of structures in the input structure file divided by arg2 of `NPRC`.
	If the total number of jobs is significantly larger than some small multiple of the number of hosts specified by arg1 of `NPRC`, load balancing can take place by sending more jobs to the machines that finished their assigned jobs faster. The number of jobs should not be too large, as this would result in overhead associated with starting many jobs and organizing the results from many jobs.
0	For non-serial searches, set the number of steps per job to the total number of steps for the search divided by the number of processors (arg1).
	For serial searches, set the number of structures to process per job to the number of structures in the input structure file divided by the number of processors.
	In either case, this setting is sub-optimal since no load balancing can be carried out.
arg3	Sleep time (sec)
	The time in seconds that the “master” MacroModel process will sleep before checking on the progress of the “slave” jobs.
0	60 seconds. This is a reasonable value.
arg4	Energy-test control
0	Perform no energy tests.
1	Perform energy tests; this is highly recommended.
	Prior to starting the distributed job, an energy calculation on the starting structure is performed on each processor. The resulting energies are compared. If any value differs from that on the current processor by more than 0.001 kJ/mol or 0.1% (whichever is smaller), the user is notified and the job is terminated. This is to make sure that the version of MacroModel and the associated solvent and force-field files are the same on all hosts.

`SPAT` — SPecial Atom Treatment

This opcode is used to instruct MacroModel how to proceed when it encounters certain atom types.

arg1-4	Atom types to be flagged
	Specify which atom types are flagged by MacroModel for special treatment. See MacroModel Atom Types for a listing of MacroModel atom types.
arg5	Action to be taken on flagged atom types
1	Instructs MacroModel to stop when it encounters any of the atom types specified in args 1-4.
	Multiple `SPAT` lines may be used, and arg5 may differ on each of them (it applies only to the atom types given in the same `SPAT` line).

Miscellaneous and Debugging Opcodes

DEBG — DEBuGging

DUMP — DUMP connection table

GEOM — obtain GEOMetric information about the molecule

TIME — report cpu TIME (user+system)

JWRT — Journal WRiTe

JRED — Journal REaD

NPRC — Number of PRoCessors (distributed MacroModel calculations)

SPAT — SPecial Atom Treatment