Print

Print


Eleanor,

I’ve come up with a work-around for both problems. The problem with the tRNA file, I countered by using that pdb file as the fixed file, and so got round the need to write out a rotated pdb file (it still did not write out the output file, even though the rmsd was 0!).

The second problem, with the Stau coordinates, I got round by rotating them first into a new orientation. That suggests that there is a mathematical singular point for the rotation problem. I don’t know whether or not that can be fixed by some test on the singularity of the rotation matrix or something like that.

Many thanks for your shoulders while I worked out of this rabbit hole!

Charlie

On Sep 10, 2018, at 8:12 AM, Eleanor Dodson <[log in to unmask]<mailto:[log in to unmask]>> wrote:

yes - I see that but when I run lsqkab on your two coordinate sets   lsqkab works OK.

Here is the command file  - i made a guess on what to match..


*******************************************************************
/tmp/shiftField_15_1_com.tmp
*******************************************************************
 title [No title given]
fit res CA 11 to 100 -
    chain A
match 8 to 97 -
    chain A
fit res CA 101 to 130 -
    chain A
match 100 to 129 -
    chain A
output -
    xyz
end
## This script run with the command   ##########
# /y/programs/xtal/64bit/ccp4-7.0/2016-08-23-1/ccp4-7.0/bin/lsqkab XYZIN2 "/y/people/ccp4/Downloads/Ecoli_HSS_rot.pdb" XYZIN1 "/y/people/ccp4/Downloads/Stau__urz_rot.pdb" XYZOUT "/y/people/ccp4/projects/shiftField/Ecoli_HSS_rot_lsq1.pdb"
################################################

And I attach the log file.
Indeed the Ecoli one is now rotated by ~ 90 degrees and fits quite well.. There is a lot of unmatched stuff of course for the Ecoli one but that shouldnt prevent the fit..


Eleanor


On 10 September 2018 at 12:53, Eleanor Dodson <[log in to unmask]<mailto:[log in to unmask]>> wrote:
Hmm - when I try to read Therm_tRNA into the CCP4I2 GUI it says


Cordinate file failed vaildation test!
Not sure what that test is but will ask around..
e

On 10 September 2018 at 12:15, Carter, Charlie <[log in to unmask]<mailto:[log in to unmask]>> wrote:
Eleanor,

I’ll check into the tRNA TER cards. I have probably experienced difficulties with them before, so that may solve one of the problems, but not, I think, the other.

Here are two pdb files from the pymol session. Ecoli_HSS_rot.pdb is the target structure; Stau_urz_rot.pdb is the result output for the input Stau_urz.pdb file. As you can see, even though the residues are, as far as I can tell correctly indicated, the rotated file is off by about 90 degrees.

Many thanks for any help.

Charlie



On Sep 10, 2018, at 5:36 AM, Eleanor Dodson <[log in to unmask]<mailto:[log in to unmask]>> wrote:

Charlie - I will try to check this, but cant read your pse file - can you find another format?

Thanks Eleanor

On 8 September 2018 at 12:28, Carter, Charlie <[log in to unmask]<mailto:[log in to unmask]>> wrote:
Hi folks,

I continue to use lsqkab because it offers unique flexibility and features appropriate to what I do.

Currently, I’ve a project that requires such superposition and for that I created from the original pdb files a set of monomeric pdb files that have already been superposed by POSA. The POSA alignment used all the residues, and I want to orient only a subset that is structurally invariant, in order to identify possible interdomain motions.

There are ten such files; eight appear to work properly and have aligned. The other two give one or another of two pathologies. The first appears simply to get confused and output a rotated coordinate file that appears to be ~90 degrees off, despite the fact that the residue numbering is correct (see attached .pse file in which three _Urz coordinate sets are illustrated, which were superimposed on Ecoli_Urz. The Ttherm_H file is correctly rotated; the Stau_LF file is not.).

I decided to try to rotate the two files into an intermediate orientation that gave a satisfactory result, i.e. Ttherm_H. The first attempt at this involved a second problematic file, Ttherm_tRNA. LSQKSB gave the following error message, which instructs me to contact the developers. The problematic pdb file, Ttherm_tRNA.pdb, is also attached.


 *** RWBROOK error: point code unit    function
 ***                    1 -102    2    MMDB_F_Atom
 *** file   : Ttherm_tRNA_rot.pdb
 *** reason : internal error #2 -- report to developer
 *** Execution stopped.

I should note that I am still using the following version of the software:

source  /usr/local/bin/ccp4-7.0/include/ccp4.setup

The commands in my script are:

#source  /usr/local/bin/ccp4-7.0/include/ccp4.setup
lsqkab XYZINM $1 XYZINF $2 XYZOUT $1_rot.pdb << END-lsqkab
FIT RESIDUE MAIN 8 TO 82
MATCH 10 to 84
FIT RESIDUE MAIN 83 TO 111
MATCH 85 to 113
FIT RESIDUE MAIN 122 TO 129
MATCH 124 to 131
OUTPUT XYZ
END
END-lsqkab

Many thanks in advance for any help.







<15_superpose.log>


########################################################################

To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1