Previous Thread
Next Thread
Print Thread
Page 2 of 2 1 2
Re: Calculate the correlation between the original B-f
lennart #24029 04/15/10 09:12 PM
Joined: Sep 2003
Posts: 8,500
rmv Online Content
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 8,500
To use rmsf-residue.str, one needs a CHARMM coordinate set with the isotropic fluctuations in the weighting array. That stream file then computes the per-residue average, which may or may not be the best way to compare to crystal B-factors. (For optimal comparison, a simulation in the full crystal environment followed by the use of COOR DYNA might be needed. A simulation in vacuum or bulk solvent would not be expected to give identical results.)

Writing a normal mode trajectory is probably easier, something like

OPEN UNIT 3 WRITE FILE NAME nmode.trj
WRITE TRAJ PHASE 2 TEMP 298. UNIT 3

Then you can use COOR DYNA, to get a coordinate set in the format expected by the rmsf-residue.str script. Or you could compare each heavy atom directly, in which case you don't need the rmsf-residue.str script at all. You have to make these decisions, possibly with some help from people in your lab.

You must first understand for yourself what happens when you import a PDB file into CHARMM, and work out the discrepancy in the number of heavy atoms.

You must also obtain a normal mode result where the first 6 modes are zero (or nearly so). Note also that normal modes are computed in a vacuum environment (not a crystal).


Rick Venable
computational chemist

Re: Calculate the correlation between the original B-f
rmv #24148 04/27/10 09:03 PM
Joined: Oct 2009
Posts: 22
J
Jun Offline OP
Forum Member
OP Offline
Forum Member
J
Joined: Oct 2009
Posts: 22
Thank you Lennart and Rick.

Now I want to get a second derivative matrix.

I used the following input file

vibran nmod 528
diag
fluctuations atom
open unit 11 write form name second.dat
write second card unit 11
close unit 11
end


Then I have the following second.dat


176
-259.2751355843
-0.0127387887 -1.3915580533 -1.8785803856
-1.3363841164 -0.2851400195 -1.0916435796
-0.9202203323 1.3173123253 0.8601555325
-1.3926104184 -0.3717030631 -0.6271204575
-0.0269574712 -0.7133016063 -1.0950389883
-0.5542056111 0.2874617640 -0.2533952282
0.3986226976 -0.7161613857 0.2624511014
-0.3747245064 -0.0635467815 0.0554824838
-0.4796916714 -0.0156409026 -0.0313864651
2.5423632286 -0.1014721861 1.6924151958
-0.3699589043 -0.1764678802 -0.0902327294
-0.3855055013 -0.1967257332 0.0133131357
-3.5402089878 -0.2266930691 -2.5701247286
2.3105129531 1.8582760611 2.8243669254
4.5954606477 0.0526673184 1.5340555076
-4.2013243560 -0.3288930107 -1.6032829122
1.4589680971 0.5293008264 1.1754925901
2.7954168671 0.1766424640 2.0483252335
-0.4258072036 -0.2580591406 -0.8128065411
-1.3951360022 -0.6040137089 -0.3911179645
.......
.......
.......
-11.9621063155
3.2407886592
-2.2795395094
-0.8377439603
0.3767962156
-0.0199160833
0.1308252870
0.0443599115
-0.1396421016
-0.8579321238
-1.7997533082
0.6597106494
-2.7491882344
0.1745593803
-0.0775330576
0.2375622414
-0.0410735763
-0.0191195431
0.2189660868
-29.1638083988
-15.4716628259
-26.5431404431
-4.4382348314
-29.9792622641
-35.4522100631
-3.9091947595
-2.9579707332
........
........
........
1.3516100000 0.7836300000 -1.9970800000
0.0290600000 -0.1131400000 -0.7275600000
-0.5088100000 0.4509100000 0.0218000000
-0.7850300000 -1.3442900000 -1.2051900000
-0.1353400000 -1.9744900000 -1.8460000000
-1.6067200000 -0.9787200000 -1.8506300000
-1.5234200000 -2.3565300000 0.1102000000
1.3656800000 -0.5959800000 -0.1182700000
2.2029700000 -1.1436000000 -0.8911400000
1.5697700000 -0.4066400000 1.1069300000


My questions is

1) Why the matrix look like this ? How do I get n by n matrix(square) ? I tried to find a similar example in this forum, however I couldn't.


Thank you.

Re: Calculate the correlation between the original B-f
Jun #24150 04/27/10 09:09 PM
Joined: Sep 2003
Posts: 8,500
rmv Online Content
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 8,500
You'll probably have to either look at the source code, or reverse engineer the format via some small molecule tests (or both).


Rick Venable
computational chemist

Re: Calculate the correlation between the original B-f
rmv #25478 09/27/10 08:49 PM
Joined: Oct 2009
Posts: 22
J
Jun Offline OP
Forum Member
OP Offline
Forum Member
J
Joined: Oct 2009
Posts: 22
When I check the source code, I found this.

IF(PRLEV.LT.0) RETURN
WRITE(IUNSCD,725) NATOM
725 FORMAT(I5)
WRITE(IUNSCD,726) EPROP(EPOT)
726 FORMAT(3F20.10)
WRITE(IUNSCD,726) (DX(K),DY(K),DZ(K),K=1,NATOM)
WRITE(IUNSCD,727) (DD1(K),K=1,NIJ)
WRITE(IUNSCD,726) (X(K),Y(K),Z(K),K=1,NATOM)
727 FORMAT(F20.10)



I only want to extract the middle part, (DD1(K),K-1,NIJ) not top or bottom part. Is the best way modifying the source code? Is there any other way to extract the values?

If I need to erase two lines, WRITE(IUNSCD,726) (DX(K),DY(K),DZ(K),K=1,NATOM) & WRITE(IUNSCD,726) (X(K),Y(K),Z(K),K=1,NATOM) how do I compile the modified code? I am a beginner in unix, too. Thank you.

Page 2 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: 0.012s Queries: 22 (0.005s) Memory: 0.9308 MB (Peak: 1.0125 MB) Data Comp: Off Server Time: 2020-10-01 23:23:26 UTC
Valid HTML 5 and Valid CSS