| CFF |
The Getting started section is designed to step a first-time user of the Insight II program user through a CFF calculation. It does, however, assume you have gained some familiarity with the Insight interface. If not, please see the Insight documentation before continuing with this section.
Users already familiar with Insight II
If you are already familiar with the Insight II program, please just scan the provided .log files and note that the Assign_CFF command replaces both the Select and Potentials commands of the Forcefield pulldown. Use of default values leads to the released CFF. The Sample computation using CFF section describes what should be done when parameters are missing.
To demonstrate the use of the CFF forcefield we will step through the basic, but essential, steps with a urea model.
Getting started
In most applications, addition of new parameters should not be necessary; therefore, this section is optional.
Sample computation using CFF
This section provides instructions for assigning the CFF forcefield, adding parameters to the .frc forcefield file, and running a simple minimization with the Discover program. Under Checking for undefined parameters and atom types, an example is presented where silicon-related parameters were missing for the model MPS. On the following pages we show how to identify missing parameters, how to supply new ones, and how to use the new data file to run a minimization.
Where to look for missing parameters
If a potential type is assigned but not present in the .frc forcefield file the Discover program does not run, as we see when attempting to run a minimization on MPS after execution of Assign_CFF. Here are the steps taken, which if all parameters were present, would have resulted in a successful minimization using CFF: #Running: /ax10/release970/irix5r4/biosym_exe/insightII
m: Builder
m:Restore_Folder chapt3.psv Unique *
m:Assign_CFF Fix Fix Accept MPS
m:Discover
m:Minimize -List VA09A 100 0.001 Charges Cross -Morse
m:Run Local -List MPS Interactive -Strategy Explicit Run_Minimization -Run_Dynamics -PBC Reduce_Output
We begin by restoring a folder from a previous session, in this case from the example you made in the Getting started section , which was saved using the File/Save_Folder command. Next, assign the CFF forcefield with the Forcefield/Assign_CFF command. You can obtain a model from a .car/.mdf file, from a fragment library, or from the sketcher. At this point we are first warned of a problem, since the following message is written to the textport.
(Executing: Assign_CFF Fix Fix Accept MPS) atypq: missing bond increment for si,osh, zero used atypq: missing bond increment for osh,hos, zero usedEven though this was not at first noticed, when the request for the Discover module was made, instead of seeing the menu bar appear, a Potentials parameter block appeared, indicating that the Insight II potentials_ok flag had not yet been set. For this example, Cancel was selected and the Discover minimization was started. On writing the .car and .mdf files another warning is given:
Atom with unknown potential type found. Topology file not usable as simulation inputand soon after, notice is given that the minimization failed.
In each case, we recommend deriving the parameters from ab initio methods described elsewhere in this document, by contacting one of the authors, or from the publication of your choice.
Parameters presented in this tutorial are for example only and have not yet been derived by MSI's Potential Energy Functions Consortium.
Create a new directory and cd to it using the commands:
> mkdir cff_mod > cd !$Now make a copy of the cff_parameter_template file using the command:
> cp $BIOSYM_LIBRARY/cff/cff_parameter_template .This file has lines resembling those of the $BIOSYM_LIBRARY/cff91.frc file and is provided to allow correct entry of parameter types and error checks on input data.
Essential are the "#atom_types cff" and "#equivalence cff" data; for the others, the requirements depend on the system. Enter your parameters into the cff_parameter_template file. The result should look like the following:
Note, for CFF version 3.3 is required.
#atom_types cff
!Ver Ref Type Mass Element Connection Comment
!---- --- ---- ---------- ---------- ---- -------------------------------------------------
3.3 1 si 28.08601 Si 4 silicon atom
3.3 10 osh 15.99490 O 2 oxygen atom in terminal hydroxyl group on silicon
3.3 10 hos 1.007825 H 1 hydrogen atom in terminal hydroxyl group on silicon
#equivalence cff
! Equivalences
! ----------------------------------------------------------------------
!Ver Ref Type NonB Bond Angle Torsion OOP
!--- ----- ------ -------- ------- ------- --------- --------
3.3 1 si si si si si si
3.3 10 osh osh osh osh osh osh
3.3 10 hos hos hos hos hos hos
#auto_equivalence cff_auto
#quartic_bond cff
> E = K2 * (R - R0)^2 + K3 * (R - R0)^3 + K4 * (R - R0)^4
!Ver Ref I J R0 K2 K3 K4
!---- --- ---- ---- ----------- --------------- ---------------- --------------
3.3 9 si c 1.89966 189.60364 -279.22096 307.41352
3.3 9 si h 1.47926 202.76983 -305.35026 280.25846
3.3 9 cp si 1.86139 233.34329 -276.96923 161.86592
3.3 10 si osh 1.6225 420.1240 -845.4110 1438.5300
3.3 10 osh hos 0.9497 702.6730 -1531.7700 1814.7300
#quartic_angle cff
> Delta = Theta - Theta0
> E = K2 * Delta^2 + K3 * Delta^3 + K4 * Delta^4
!Ver Ref I J K Theta0 K2 K3 K4
!---- --- ---- ---- ---- --------------- ------------- -------------- -------------
3.3 9 cp cp si 120.04219 30.47886 -23.5939 .03167
3.3 9 cp si h 109.49316 41.94968 -42.46593 48.25418
3.3 9 si c h 112.04552 28.76214 -13.94228 .00000
3.3 10 si osh hos 123.5920 14.9634 -17.3643 57.9495
.
.
.
#torsion_3 cff
> E = SUM(n=1,3) { V(n) * [ 1 - cos(n*Phi - Phi0(n)) ] }
!Ver Ref I J K L V1 Phi0 V2 Phi0 V3 Phi0
!---- -- --- --- --- --- ----------- ----- ----------- ----- ------------ -----
3.3 9 h si c h 0.00000 0.0 0.00000 0.0 -0.06666 0.0
3.3 9 cp cp si h 0.00000 0.0 0.00000 0.0 -0.04311 0.0
3.3 9 cp cp si c 0.00000 0.0 0.00000 0.0 -0.03311 0.0
3.3 9 si cp cp cp 0.00000 0.0 4.42702 0.0 0.00000 0.0
3.3 9 si cp cp h 0.00000 0.0 1.40934 0.0 0.00000 0.0
3.3 9 cp si osh hos 0.00000 0.0 0.00000 0.0 -0.06569 0.0
3.3 1 h si osh hos 0.1863 0.0 -0.4338 0.0 -0.2121 0.0
3.3 1 c si osh hos -0.6732 0.0 -0.4778 0.0 -0.1670 0.0
.
.
.
#wilson_out_of_plane cff
> E = K * (Chi - Chi0)^2
!Ver Ref I J K L K Chi0
!--- --- ---- ---- ---- ---- ----------- ---------
3.3 1 si cp cp cp 7.1794 0.0000
.
.
.
#nonbond(9-6) cff
> E = eps(ij) [2(r(ij)*/r(ij))**9 - 3(r(ij)*/r(ij))**6]
> where r(ij) = [(r(i)**6 + r(j)**6))/2]**(1/6)
>
> eps(ij) = 2 sqrt(eps(i) * eps(j)) *
> r(i)^3 * r(j)^3/[r(i)^6 + r(j)^6]
@combination sixth-power
@type r-eps
!Ver Ref I r eps
!---- --- ---- --------- ---------
3.3 9 si 4.5532 0.1963
3.3 10 osh 3.4518 0.1581
3.3 10 hos 2.3441 0.0998
#bond_increments cff
!Ver Ref I J DeltaIJ DeltaJI
!--- --- ---- ---- ----------- ----------
3.3 10 osh si -0.1333 0.1333
3.3 10 hos osh 0.0651 -0.0651
#end
A few sections have been left out, and at this point we do not yet know if we have provided all necessary parameters, but with this information, the Discover program should be able to perform a minimization.After all new parameters are entered in the template file, they are entered into the encoded .frc forcefield file when you enter:
> add_to_cffThis is a program that is provided with the CFF forcefield. It first reads the cff_parameter_template file, looking for repeated data, incorrect entries and other kinds of errors and, if everything looks alright, it reads the encoded $BIOSYM_LIBRARY/cff/cff.frc forcefield file and writes into the current directory a new encoded file with the new parameters inserted into the appropriate place. This file is named cff_mod.frc and the corresponding binary file, cff_mod.bin should also be present. If no error messages is printed, just move this file to the location of the original using the commands:
> mv cff_mod.frc $BIOSYM_LIBRARY/cff/cff.frc > mv cff_mod.bin $BIOSYM_LIBRARY/cff/cff.binand you may return to the Insight II session.
#Running: /ax10/bioroot/develop/release960/irix5r3/biosym_exe/insightII #Log Created: Wed Oct 18 16:45:53 1996 #Insight II Version 97.0 - Molecular Modeling System m:Restore_Folder ../chapt3.psv Unique * m:Label Molecule Single_property Atom On MPS Potential m:Assign_CFF Fix Fix Accept MPS m:Discover m:Minimize -List VA09A 100 0.001 Charges Cross -Morse m:Run Local -List MPS Interactive -Strategy Add_Auto Run_Minimization -Run_Dynamics -PBC Reduce_OutputThis time Discover minimizes the model using the modified CFF forcefield. In the next section we look at the Discover output and determine if the results are acceptable.
The mps0.out file describes what the Discover program did. The name of the forcefield, a summary of what was to be done, and finally, the initial and final energy and derivative results are given:
DISCOVER Molecular Simulation Program, Version 2.97
Date: 18-Oct-96 Time: 16:49:02
********************************************************************************************
*Line 1 ! INPUT FILE FOR DISCOVER GENERATED BY INSIGHT
*Line 2 !
*Line 3 !
********************************************************************************************
*Line 4 overlap = 0.01
********************************************************************************************
********************************************************************************************
*Line 5 begin simulation
*Line 6 * add-automatic bond torsion valence out-of-plane
********************************************************************************************
Force Field parameters will be selected from the force field library.
The Binary Force Field used is :
Confirm that the correct
forcefield file was used
/net/iris78/usr/people/wty/BIOSYM_LIBRARY//cff/cff.bin
The Binary File was generated from a text file containing the following version information:
cff.frc 3.3 15-May-96
FORCE FIELD INPUT OPTIONS
=========================
potential data options
default updated
cutoff 0.10E+33 0.10E+33 non-bond neighbor list - all residues within cutoff distance are included
cutdis 0.10E+33 0.10E+33 non-bond energy cutoff distance - interactions greater than this are neglected
swtdis 0.00E+00 0.00E+00 switch distance the distance over which the switching function reduces the non-bond energy to zero
Cross term energies will be included.
A harmonic potential will be used for bond energies.
non-bond switching function: lower bound 0.100000E+07 upper bound 0.100000E+07
coordinate file title: input file for discover 0.0000000E+00
---------------------- !DATE Wed Oct 18 16:48:48 1995
for molecule 1 no. of residues read= 1 no. of atoms read 19
residue names:
MPS 1
END OF INPUT: no. of molecules read 1
total no. atoms 19 limit 25000
total no. groups 9 limit 8000
total no. residues 1 limit 7500
total no. molecules 1 limit 7500
total no. bonds 19 limit 30000
total no. valence angles 3 limit 50000
total no. torsion angles 42 limit 75000
total no. out of planes 6 limit 7500
total no. theta*theta angles 42 limit 87500
total no. excluded atoms 7 limi 325000
3 groups and 0 residues within molecule 1 are Not Neutral.
The Total Charge of the molecule is 0.00000
The following protocol will be used for parameter assignments:
For this calculation automatic parameters were
allowed, typically they
would not be.
|----------------------------------------------------------------------------------------------------------| | type non-automatically 1st non-automatically automatically | only then automatically only |----------------------------------------------------------------------------------------------------------| | | bond STOP | valence STOP | torsion STOP | out-of-plane STOP | cross-term CONT |----------------------------------------------------------------------------------------------------------| | If "STOP" is specified, the program will stop when parameters for an | internal are not found in the library. If "CONT" is specified, the | program will select a value of zero for the parameter and continue. |----------------------------------------------------------------------------------------------------------|The following table summarizes the number of interactions for which parameters were either found explicitly in the force field file, automatically generated from bond order information, or set to zero if not available. Each of the major classes of interactions is reported separately. A detailed listing of all the interactions requiring automatically generated parameters has been printed in a separate file. The default name of the file is "fname".prm (logical unit 1).
+------------------------------------------------------------------------------------------------------------------------+ | BOND VALENCE TORSION OUT-OF- ANGLE*ANGLE | LENGTHS ANGLES ANGLES PLANES CROSS TERMS |Most parameters have been explicitly assigned.
| Available explicitly 19 26 34 6 18
| Assigned automatically 0 5 8 0 0
| Set to zero 0 0 0 0 24
+-------------------------------------------------------------------------------------------------------------------------+
***************************************************************************************************
Warning: 13 Interaction Parameters were Assigned Automatically.
A detailed listing of the parameter assignments resides in the
.prm file
***************************************************************************************************
***************************************************************************************************
Warning: 24 Interaction Parameters were set to Zero.
A detailed listing of the parameter assignments resides in the
.prm file.
***************************************************************************************************
total time for system input and connectivity generation 0.22487 secs
Input file assignments:
--------------------------------
File type
coordinate -- assigned externally.
molecular data -- assigned externally.
force field -- assigned externally.
restart -- assigned externally.
template -- assigned externally.
The random number seed is 24917.
------------------------------------------------------
********************************************************************************************
*Line 7 reduce
*Line 8 !
********************************************************************************************
*Line 9 set dielectric = 1.000000
********************************************************************************************
A scalar dielectric constant of 1.0000 is being used
********************************************************************************************
*Line 10 !
*Line 11 !
*Line 12 !
********************************************************************************************
*Line 13 Minimize
*Line 14 * no morse
*Line 15 * for 100 iterations
*Line 16 * using VA09A
*Line 17 * until the maximum derivative is less than 0.001000000 kcal/A
********************************************************************************************
MINIMIZATION
============
Minimizer selected: BFGS.
Number of steps: 100
Maximum derivative: 0.100E-02
A harmonic potential will be used for bond energies.
Cross term energies will be included.
Status Before Minimization
==========================
Initial Energies
---------------------
total energy components for: input file for discover
12.465074 kcal=total energy
15.833988 kcal=bond energy
4.812988 kcal=theta energy
-9.588978 kcal=phi energy
0.000000 kcal=out of plane energy
2.330275 kcal=bond*bond energy
0.234109 kcal=bond*theta energy
0.012386 kcal=theta*theta energy
-9.738968 kcal=bond*phi energy
-0.503727 kcal=theta*phi energy
-0.004803 kcal=theta*theta*phi energy
-1.000405 kcal=bond*bond (1-3) energy
0.000000 kcal=op*op energy
0.000000 kcal=phi*phi energy
0.000000 kcal=hydrogen-bond energy
4.752587 kcal=nonbond energy
19.972122 kcal=non-bond repulsion energy
-15.219535 kcal=non-bond dispersion energy
5.325622 kcal=coulomb energy
0.000000 kcal=forcing potential; this contribution
has been subtracted from the total energy
Initial Derivative Summary
----------------------------------------
Atom Residue Molecule Derivative (kcal/mole-Angstrom)
Name Name No. Number dv/dx dv/dy dv/dz
------- -------- ---- -------------- ------------------ ---------------
Min: H62 MPS 1 1 -2.431803 -0.968104 0.762886
Max: O MPS 1 1 64.652622 -136.480528 29.821069
Average absolute derivative 18.5469776
Stand. dev. of abs. deriv. 23.5798846
RMS derivative 30.0000223
Starting minimization using BFGS
ABNORMAL RETURN FROM MINIMIZATION
Note, have not yet
reached convergence.
No convergence after 100 iterations
The maximum derivative criteria used was: 0.00100
The maximum derivative achieved is: 1.11654
100 iterations took 131 energy evaluations and 1.670 secs.
Time per energy evaluation is 0.013
Status After Minimization
===================
Final Energies
--------------------
total energy components for: input file for discover
Total energy, however is
decreasing
-7.283994 kcal=total energy
3.762287 kcal=bond energy
1.031829 kcal=theta energy
-9.969428 kcal=phi energy
0.000317 kcal=out of plane energy
0.931849 kcal=bond*bond energy
0.102982 kcal=bond*theta energy
-0.006119 kcal=theta*theta energy
-10.138668 kcal=bond*phi energy
-0.866683 kcal=theta*phi energy
-0.031910 kcal=theta*theta*phi energy
-0.327998 kcal=bond*bond (1-3) energy
0.000000 kcal=op*op energy
0.000000 kcal=phi*phi energy
0.000000 kcal=hydrogen-bond energy
3.254335 kcal=nonbond energy
16.319635 kcal=non-bond repulsion energy
-13.065300 kcal=non-bond dispersion energy
4.973213 kcal=coulomb energy
0.000000 kcal=forcing potential; this contribution
has been subtracted from the total energy
Final Derivative Summary
------------------------
Atom Residue Molecule Derivative (kcal/mole-Angstrom)
Name Name No. Number dv/dx dv/dy dv/dz
----- ------- ------- ---------------- ------------------- --------------
Min: C MPS 1 1 0.006577 0.077554 0.042398
Max: O MPS 1 1 -0.327573 -1.116535 -0.008692
Average absolute derivative 0.2119085
Stand. dev. of abs. deriv. 0.2065751
RMS derivative 0.2959367
********************************************************************************************
*Line 18 !
*Line 19 !
********************************************************************************************
Total time used by DISCOVER: 2.04 second(s).
DISCOVER completed: Date: 18-Oct-96 Time: 16:49:36
The mps0.prm file describes how many parameters were available from the .frc forcefield file, how many were automatically assigned, and how many were set to zero. When bond, angle, or torsion parameters are set to zero, the Discover program should not run, but reasonable results may be obtained if other parameter types are set to zero. It is always important to check this file to see what may have been automatically assigned, since if they involve regions you are interested it would be better to have them derived.
Parameter Assignment |==================================================================================| | BONDS: |-------------------------------------------------------------------------------------------------------|All bond parameters are avaiable
| 19 bond parameters were assigned non-automatically | 0 bond parameters were assigned automatically | 0 bonds had parameters set to zero as the interaction was not in the library. |-------------------------------------------------------------------------------------------------------| |------------------------------------------------------------------------------------------------------------------------------| | VALENCE ANGLES: | automatic parameter assignments |------------------------------------------------------------------------------------------------------------------------------| | Atom | Atom | Actual | Automatic | Force Constants | Number | Name | Interaction | Interaction | * = not found and set to zero |------------------------------------------------------------------------------------------------------------------------------| | 11 12 14 | C5 SI C6 | cp si c | *7 si_ c_ | 44.40 113.50 0.00 0.00 0.00 0.00 0.00 | 11 12 18 | C5 SI O | cp si osh | *7 si_ c_ | 44.40 113.50 0.00 0.00 0.00 0.00 0.00 | 13 12 14 | HSI SI C6 | h si c | *2 si_ h_ | 31.80 112.00 0.00 0.00 0.00 0.00 0.00 | 13 12 18 | HSI SI O | h si osh | *2 si_ h_ | 31.80 112.00 0.00 0.00 0.00 0.00 0.00 | 14 12 18 | C6 SI O | c si osh | *7 si_ c_ | 44.40 113.50 0.00 0.00 0.00 0.00 0.00 |------------------------------------------------------------------------------------------------------------------------------|However several angle parameters are still missing.
| 26 theta parameters were assigned non-automatically | 5 theta parameters were assigned automatically | 0 valence angles had parameters set to zero, as the interaction was not found in the library. |---------------------------------------------------------------------------------------------------------------------------------| |---------------------------------------------------------------------------------------------------------------------------------| | TORSION ANGLES: | automatic parameter assignments |---------------------------------------------------------------------------------------------------------------------------------| | Atom | Atom | Actual | Automatic | Force Constants | Number | Name | Interaction | Interaction | * = not found and set to zero |---------------------------------------------------------------------------------------------------------------------------------| | 1 11 12 18 | C C5 SI O | cp cp si osh | * cp_ si_ * | 0.00 0.00 0.17 0.00 | 9 11 12 18 | C4 C5 SI O | cp cp si osh | * cp_ si_ * | 0.00 0.00 0.17 0.00 | 15 14 12 11 | H61 C6 SI C5 | h c si cp | * c_ si_ * | 0.00 0.00 0.11 0.00 | 16 14 12 11 | H62 C6 SI C5 | h c si cp | * c_ si_ * | 0.00 0.00 0.11 0.00 | 17 14 12 11 | H63 C6 SI C5 | h c si cp | * c_ si_ * | 0.00 0.00 0.11 0.00 | 15 14 12 18 | H61 C6 SI O | h c si osh | * c_ si_ * | 0.00 0.00 0.11 0.00 | 16 14 12 18 | H62 C6 SI O | h c si osh | * c_ si_ * | 0.00 0.00 0.11 0.00 | 17 14 12 18 | H63 C6 SI O | h c si osh | * c_ si_ * | 0.00 0.00 0.11 0.00 |---------------------------------------------------------------------------------------------------------------------------------|As are some torsion parameters
| 34 torsion parameters were assigned non-automatically. | 8 torsion parameters were assigned automatically. | 0 torsion angles had parameters set to zero, as the interaction was not found in the library. |---------------------------------------------------------------------------------------------------------------------------------| | OUT-OF-PLANES: |---------------------------------------------------------------------------------------------------------------------------------| | 6 out-of-plane parameters were assigned non-automatically | 0 out-of-plane parameters were assigned automatically | 0 out-of-planes had parameters set to zero, as the interaction was not found in the library. |---------------------------------------------------------------------------------------------------------------------------------| | THETA*THETA CROSS TERMS: | 9 12 1 11 | C4 SI C C5 | cp si cp cp | * 0.00 | 1 12 9 11 | C SI C4 C5 | cp si cp cp | * 0.00 | 1 9 12 11 | C C4 SI C5 | cp cp si cp | * 0.00 | 13 14 11 12 | HSI C6 C5 SI | h c cp si | * 0.00 | 13 18 11 12 | HSI O C5 SI | h osh cp si | * 0.00 | 11 14 13 12 | C5 C6 HSI SI | cp c h si | * 0.00 | 11 18 13 12 | C5 O HSI SI | cp osh h si | * 0.00 | 14 18 11 12 | C6 O C5 SI | c osh cp si | * 0.00 | 11 13 14 12 | C5 HSI C6 SI | cp h c si | * 0.00 | 11 18 14 12 | C5 O C6 SI | cp osh c si | * 0.00 | 11 13 18 12 | C5 HSI O SI | cp h osh si | * 0.00 | 11 14 18 12 | C5 C6 O SI | cp c osh si | * 0.00 | 14 18 13 12 | C6 O HSI SI | c osh h si | * 0.00 | 13 18 14 12 | HSI O C6 SI | h osh c si | * 0.00 | 13 14 18 12 | HSI C6 O SI | h c osh si | * 0.00 | 15 16 12 14 | H61 H62 SI C6 | h h si c | * 0.00 | 15 17 12 14 | H61 H63 SI C6 | h h si c | * 0.00 | 12 16 15 14 | SI H62 H61 C6 | si h h c | * 0.00 | 12 17 15 14 | SI H63 H61 C6 | si h h c | * 0.00 | 16 17 12 14 | H62 H63 SI C6 | h h si c | * 0.00 | 12 15 16 14 | SI H61 H62 C6 | si h h c | * 0.00 | 12 17 16 14 | SI H63 H62 C6 | si h h c | * 0.00 | 12 15 17 14 | SI H61 H63 C6 | si h h c | * 0.00 | 12 16 17 14 | SI H62 H63 C6 | si h h c | * 0.00 |------------------------------------------------------------------------------------------------------|For now, setting the cross terms to zero is acceptable.
| 18 theta*theta parameter(s) assigned non-automatically | 0 theta*theta parameter(s) assigned automatically | 24 theta*theta(s) had parameters set to zero, as the interaction was not found in the library. |==================================================================================|