Previous Thread
Next Thread
Print Thread
Page 1 of 2 1 2
one problem about box setting
#5738 03/09/05 03:38 AM
Joined: Jan 2004
Posts: 60
H
Forum Member
OP Offline
Forum Member
H
Joined: Jan 2004
Posts: 60
Dear all,
I want to set a box without PBC, just only a box. My purpose is I just want my molecule to move in the box. Here is my setting,
...........
set 7 130.000
set 8 300.000
set 9 130.000

crystal define monoclinic @7 @8 @9 90.0 60.0 90.0
crystal build cutoff 15.0

............

But the program can't run, and if I delete the line "crystal build....", it becomes to run, but I am not sure my molecule moves just in the box, not outside the box.
So, what's wrong with my input file?
Thank you very much!!

hauren32

Re: one problem about box setting
hauren32 #5739 03/09/05 07:21 AM
Joined: Sep 2003
Posts: 4,794
Likes: 2
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 4,794
Likes: 2
Everything is wrong. This will not work as intended. The crystal and image facilities are only for PBC simulations, and cannot be used to implement a hard-wall box.


Lennart Nilsson
Karolinska Institutet
Stockholm, Sweden
Re: one problem about box setting
hauren32 #5740 03/09/05 08:45 AM
Joined: Jan 2004
Posts: 60
H
Forum Member
OP Offline
Forum Member
H
Joined: Jan 2004
Posts: 60
Dear sir,
Thank you for your reply.
As you mention, my setting is just for PBC, but if I want to set the hard-wall box, how should I do? or can charmm program do it?
Thank you very much~

hauren32

Re: one problem about box setting
hauren32 #5741 03/09/05 08:59 AM
Joined: Sep 2003
Posts: 4,794
Likes: 2
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 4,794
Likes: 2
you best bet is probably to construct physical walls using very hard L-J spheres (non-charged) constrained on a grid; this is not a all difficult to do.


Lennart Nilsson
Karolinska Institutet
Stockholm, Sweden
Re: one problem about box setting
hauren32 #5742 03/09/05 09:20 AM
Joined: Jan 2004
Posts: 60
H
Forum Member
OP Offline
Forum Member
H
Joined: Jan 2004
Posts: 60
Dear sir,
Thank you very much!
But could you tell more clearly? Is there any charmm document for me? or some example? I don't have any experience about this.

hauren32

Re: one problem about box setting
hauren32 #5743 03/09/05 09:55 AM
Joined: Sep 2003
Posts: 4,794
Likes: 2
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 4,794
Likes: 2
This is really very simple, but sligthly tedious, and I cannot do it for you.
The basic steps are to create a residue (in your favorite RTF) containing just one non-charged (carbon) atom and then to generate a segment with as many such residues as you need to build a wall. Coordinates can be generated manually, or with a simple program, or even inside CHARMM using a loop. Finally you may want to play with the parameters (in the parameter file) for the size (Rmin) and attractive potential depth (epsilon) of the atoms that make up your wall.


Lennart Nilsson
Karolinska Institutet
Stockholm, Sweden
Re: one problem about box setting
lennart #5744 03/09/05 01:10 PM
Joined: Jan 2004
Posts: 60
H
Forum Member
OP Offline
Forum Member
H
Joined: Jan 2004
Posts: 60
Daer sir,
Sorry, sir. Although you told very clearly, I was still confused. I don't know why I have to create a residue containing just one carbon atom, and the wall as if a carbon wall? Is there no any other method? Actually, I don't know how to deal with my work, I just want to set a monoclinic box with its length, 130x250x130, and angles, 90,60,90.
Thanks a lot!!

hauren32

Re: one problem about box setting
hauren32 #5745 03/09/05 02:44 PM
Joined: Sep 2003
Posts: 4,794
Likes: 2
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 4,794
Likes: 2
You do not have to have just one atom per residue and it does not have to be a carbon atom, all I suggest is that you construct physical walls of immovable particles that mainly provide a repulsive interaction to keep your system confined within the volume of interest.


Lennart Nilsson
Karolinska Institutet
Stockholm, Sweden
Re: one problem about box setting
hauren32 #5746 03/09/05 05:44 PM
Joined: Sep 2003
Posts: 8,499
rmv Online Content
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 8,499
First, it's not clear at all why you would wish to do this, or at least why you don't use a large cubic box. A monoclinic lattice is most often used for crystals with P21 symmetry. Exactly what are you trying to do, and why?

The very short cutoff for CRYSTAL BUILD may a problem at the very least-- it should probably be more like half the longest cell dimension, esp. if there's considerable vacuum space inside the volume.

Try posting the last page or so of the output with some error messages; "the program can't run" doesn't provide any useful information.


Rick Venable
computational chemist

Re: one problem about box setting
rmv #5747 03/09/05 10:58 PM
Joined: Sep 2003
Posts: 250
Forum Member
Offline
Forum Member
Joined: Sep 2003
Posts: 250
So this would suggest a "crystal build cutoff 150 noper" in this case?

Re: one problem about box setting
jb007 #5748 03/09/05 11:14 PM
Joined: Sep 2003
Posts: 8,499
rmv Online Content
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 8,499
Assuming the molecule isn't too small; the CUTOFF option of CRYSTAL BUILD is to setup the image transformations, and if it's too short there can be missing transformations. It is not directly related to either CUTNB or CUTIM, as often seems to be assumed. For the example given, only image transformations within 15 A of any protein atom would be defined, which for a protein in an oversized vacuum box, could be none at all.


Rick Venable
computational chemist

Re: one problem about box setting
rmv #5749 03/10/05 03:45 AM
Joined: Jan 2004
Posts: 60
H
Forum Member
OP Offline
Forum Member
H
Joined: Jan 2004
Posts: 60
Dear sir,
Thsnks for your suggestion.
Actually, my lipid layer has a tilt angle of 30 degree, and so I have to use a monoclinic box or crystal for it. Because my lipid layer is very large, I don't want to use the PBC, I just want a only one box. The former will cost me a lot of calculation time.
But if I want to use the 2D PBC, such as xy plane, what should I do?
Here are some of my input file and output data.
.................................
SET 7 130.000000
SET 8 250.000000
SET 9 130.000000

!scale @7 @8 @9
crystal define monoclinic @7 @8 @9 90.0 60.0 90.0
crystal build cutoff 150.0 noper

!BOUND RDBOUND xsize @7 ysize @8 zsize @9 CUTNB 15.000000

define mix1 sele (bynu 1:21888 .and. segid M_C1 ) END
define mix2 sele (bynu 21889:31009 .and. segid HSA_ ) END

! Standard cutoffs
NBONDS ATOM FSHIFT CDIE VDW VSHIFT -
CUTNB 13.0 CTOFNB 12.0 CTONNB 8.0 WMIN 1.5 -
CDIE EPS 2.00000 NORXN EPSEXT 78.5


CONStraint FIX SELE (segid M_C1 .and. type S*) END

inte sele mix1 end sele mix2 end

energy
.....................................

output data: and its error message
....................
CHARMM> !scale @7 @8 @9
CHARMM> crystal define monoclinic @7 @8 @9 90.0 60.0 90.0
Parameter: 7 -> "130.000000"
Parameter: 8 -> "250.000000"
Parameter: 9 -> "130.000000"
Crystal Parameters : Crystal Type = MONO
A = 130.00000 B = 250.00000 C = 130.00000
Alpha = 90.00000 Beta = 60.00000 Gamma = 90.00000

CHARMM> crystal build cutoff 150.0 noper

Range of Grid Search for Transformation 1 :
Lattice Vector A -3 TO 3
Lattice Vector B -2 TO 2
Lattice Vector C -3 TO 3


The number of transformations generated = 28


Number Symop A B C Distance

1 1 -2 0 -1 140.3115
2 1 -1 0 -1 60.4465
3 1 0 0 -1 33.1990
4 1 0 1 -1 141.2295
5 1 -2 0 0 78.4658
6 1 -1 -1 0 127.0020
7 1 -1 0 0 28.5523
8 1 -1 1 0 137.7274
9 1 0 -1 0 125.0049
10 1 0 1 0 125.0049
11 1 -2 0 1 110.9913
12 1 -1 -1 1 136.9129
13 1 -1 0 1 29.1855
14 1 0 -1 1 141.2295
15 1 0 0 1 33.1990
16 1 -2 0 2 144.9318
17 1 -1 0 2 117.9365
18 1 2 0 1 140.3115
19 1 1 0 1 60.4465
20 1 2 0 0 78.4658
21 1 1 1 0 127.0020
22 1 1 0 0 28.5523
23 1 1 -1 0 137.7274
24 1 2 0 -1 110.9913
25 1 1 1 -1 136.9129
26 1 1 0 -1 29.1855
27 1 2 0 -2 144.9318
28 1 1 0 -2 117.9365
THERE ARE NO ROTATIONS FOR THIS TRANSFORMATION SET
28 Transformations have been processed.



CHARMM> !OPEN UNIT 20 READ FORM NAME "2D_monoclinic.img"
CHARMM> !READ IMAGE CARD UNIT 20
CHARMM> !CLOSE UNIT 20
CHARMM>


CHARMM> !IMAGE BYsegment SELECT M_C1 END
CHARMM> !BOUND RDBOUND xsize @7 ysize @8 zsize @9 CUTNB 15.000000


CHARMM> define mix1 sele (bynu 1:21888 .and. segid M_C1 ) END
SELRPN> 21888 atoms have been selected out of 31009

CHARMM> define mix2 sele (bynu 21889:31009 .and. segid HSA_ ) END
SELRPN> 9121 atoms have been selected out of 31009


CHARMM> ! Standard cutoffs
CHARMM> NBONDS ATOM FSHIFT CDIE VDW VSHIFT -
CHARMM> CUTNB 13.0 CTOFNB 12.0 CTONNB 8.0 WMIN 1.5 -
CHARMM> CDIE EPS 2.00000 NORXN EPSEXT 78.5

<MKIMAT>: updating the image atom lists and remapping
Transformation Atoms Groups Residues Min-Distance
1 C001 has 0 0 0 138.20
2 C002 has 0 0 0 56.60
3 C003 has 0 0 0 22.95
4 C004 has 0 0 0 126.07
5 C005 has 0 0 0 70.24
6 C006 has 0 0 0 126.40
7 C007 has 5586 147 147 0.00
8 C008 has 0 0 0 123.82
9 C009 has 0 0 0 123.78
11 C011 has 0 0 0 59.19
12 C012 has 0 0 0 124.73
13 C013 has 0 0 0 17.12
16 C016 has 0 0 0 101.42
17 C017 has 0 0 0 115.01
Total of36595 atoms and 2977 groups and 1301 residues were included


NONBOND OPTION FLAGS:
ELEC VDW ATOMs CDIElec FSHIft VATOm VSHIft
BYGRoup NOEXtnd NOEWald
CUTNB = 13.000 CTEXNB =999.000 CTONNB = 8.000 CTOFNB = 12.000
WMIN = 1.500 WRNMXD = 0.500 E14FAC = 0.500 EPS = 2.000
NBXMOD = 5
There are 0 atom pairs and 0 atom exclusions.
There are 0 group pairs and 0 group exclusions.
<MAKINB> with mode 5 found 88766 exclusions and 81291 interactions(1-4)
<MAKGRP> found 6013 group exclusions.
<MAKGRP> found 0 image group exclusions.
Generating nonbond list with Exclusion mode = 5
== PRIMARY == SPACE FOR 7134235 ATOM PAIRS AND 0 GROUP PAIRS
NBONDA>> Maximum group spatial extent (12A) exceeded.
Size is 15.52 Angstroms and starts with atom: 13339
Please check group boundary definitions.
VEHEAP> Expanding heap size by11223040 words.
== PRIMARY == SPACE FOR 10701372 ATOM PAIRS AND 0 GROUP PAIRS
NBONDA>> Maximum group spatial extent (12A) exceeded.
Size is 15.52 Angstroms and starts with atom: 13339
Please check group boundary definitions.

General atom nonbond list generation found:
9839327 ATOM PAIRS WERE FOUND FOR ATOM LIST
138622 GROUP PAIRS REQUIRED ATOM SEARCHES

SPACE FOR 1336100 ATOM PAIRS AND 0 GROUP PAIRS

Image nonbond list generation found:
1233993 ATOM PAIRS WERE FOUND FOR ATOM LIST
0 ATOM PAIRS WERE FOUND FOR ATOM SELF LIST
2527 GROUP PAIRS REQUIRED ATOM SEARCHES


CHARMM> CONStraint FIX SELE (segid M_C1 .and. type S*) END
SELRPN> 576 atoms have been selected out of 31009

CHARMM> inte sele mix1 end sele mix2 end
SELRPN> 21888 atoms have been selected out of 31009
SELRPN> 9121 atoms have been selected out of 31009

***** LEVEL -1 WARNING FROM <CODES> *****
***** CODES> MISSING PARAMETERS
******************************************
BOMLEV

The program stopped here. I don't know what the wrong is.

Thank you very much!!


hauren32

Last edited by hauren32; 03/10/05 12:17 PM.
Re: one problem about box setting
lennart #5750 03/10/05 07:44 AM
Joined: Jan 2004
Posts: 60
H
Forum Member
OP Offline
Forum Member
H
Joined: Jan 2004
Posts: 60
Dear sir,
Thank you for your suggestion, and I will try.
But I still have a problem. I don't want my molecule moving in the box to interact with the wall. If I build the wall, will it happen?
Thank you very much!

hauren32

Re: one problem about box setting
hauren32 #5751 03/10/05 08:25 AM
Joined: Sep 2003
Posts: 4,794
Likes: 2
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 4,794
Likes: 2
???????????????????????????????

A wall with no interactions - isn't that a rather bizarre idea?

A box can be either a physical container (and then it MUST have interactions with your system) or a geometrical concept used together with periodic boundary conditions. You don't seem to want neither one nor the other.

Your "molecule moving inside the box" is rather misleading since you seem to actually have a membrane system.

It is often better to clearly state the real problem from the beginning.


Lennart Nilsson
Karolinska Institutet
Stockholm, Sweden
Re: one problem about box setting
hauren32 #5752 03/10/05 07:47 PM
Joined: Sep 2003
Posts: 8,499
rmv Online Content
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 8,499
The presence of a lipid bilayer changes things a good deal; IMHO, this should be done as a fully solvated system with PBC. There's a new method under development in CHARMM that might be a reasonable alternative, but I don't expect that to be available in a 'beta' version right away. I should add that the chain tilt does not mean you need a monoclinic lattice; we generally use a tetragonal prism for bilayer systems. Also, lipid chains are only co-aligned in the gel state, and not in biologically active fluid phase.

Finally, the error message indicates there are still problems with your initial PSF; there are missing parameters as indicated by the message

***** CODES> MISSING PARAMETERS

This has nothing to do with your crystal setup, but means you haven't yet fully described all the information needed to evaluate the energy of your model system. There are missing terms in your parameter file; you must fix that before proceeding any further. You could have saved us all some time by posting that message a lot sooner.


Rick Venable
computational chemist

Page 1 of 2 1 2

Moderated by  lennart, rmv 

Link Copied to Clipboard
Powered by UBB.threads™ PHP Forum Software 7.7.4
(Release build 20200307)
Responsive Width:

PHP: 5.6.33-0+deb8u1 Page Time: 0.018s Queries: 44 (0.010s) Memory: 1.0424 MB (Peak: 1.2139 MB) Data Comp: Off Server Time: 2020-09-30 23:31:25 UTC
Valid HTML 5 and Valid CSS