[sharp-discuss] Trivial problem?

Clemens Vonrhein vonrhein at globalphasing.com
Tue Nov 22 18:11:58 CET 2011


Hi Petr,

if you run SHARP/autoSHARP through the default (recommended) Sushi
interface (ie a browser pointing to something like
http://localhost:8080/) and not just autoSHARP through CCP4i, then
nothing in your environment has any effect whatsoever on how the server
sees or runs CCP4 programs.

The httpd is run completely unaware of any user environment (either
the user running the httpd but especially not the user connecting
through a browser to the httpd).

If you updated software relevant to SHARP/autoSHARP (mainly CCP4 and
SHELXC/D, but also ARP/wARP), make sure to check the

  /where/ever/sharp/machines/*/*.setup

files!

So my guess is that either the CCP4 version SHARP pointed to was wiped
out (renamed, moved) or maybe the OS was updated so that the still
existing binaries diodn't run any longer?

As you said: sometimes very tricky to detect.

 ----------------------------------------------------------------

My recommendation (and it seems the same way you did it) for avoiding
such problems is to install e.g. CCP4 such that each installation is
in its own directory, e.g.

  /usr/local/xtal/ccp4/6.1.3
  /usr/local/xtal/ccp4/6.1.13
  /usr/local/xtal/ccp4/6.2.0

each with their own set of setup-scripts. Like

  /usr/local/xtal/ccp4/6.1.3/setup.sh
  /usr/local/xtal/ccp4/6.1.3/setup.csh
  /usr/local/xtal/ccp4/6.1.13/setup.sh
  /usr/local/xtal/ccp4/6.1.13/setup.csh
  /usr/local/xtal/ccp4/6.2.0/setup.sh
  /usr/local/xtal/ccp4/6.2.0/setup.csh

Then I'd do once

  cd /usr/local/xtal/ccp4
  ln -s latest/setup.sh setup.sh
  ln -s latest/setup.csh setup.csh

and each time I'm happy with making a new version the default:

  cd /usr/local/xtal/ccp4
  rm latest
  ln -s 6.2.0 latest

A user just needs lines

  [ -r /usr/local/xtal/ccp4/setup.sh ] && \
    . /usr/local/xtal/ccp4/setup.sh

for sh/bash/ksh/zsh or

  if ( -e /usr/local/xtal/ccp4/setup.csh ) \
    source /usr/local/xtal/ccp4/setup.csh

for tcsh/csh in their setup files.

So a user will always get whatever the latgest default version of CCP4
is. But previous/older versions will not be overwritten and are still
available eg. forSHARP/autoSHARP ... in case one forgets to update
/where/ever/sharp/machines/*/ccp4.setup file(s).

Of course, there are probably a dozen variations on such a system
(some more likely much better). I guess the important points is to
have some kind of system (and stick to it). And I like to keep old
versions around for quite some time ... just in case.

Sorry for the slight detaour here at the end ;-)

Cheers

Clemens




On Tue, Nov 22, 2011 at 04:53:56PM +0000, Petr Leiman wrote:
> Yes, this problem drove me crazy. The exact reasons for it are unclear to me, but it appears to have been caused by updating the ccp4 suite. On all of our systems the most recent version of the ccp4 package is soft-linked as /blah/blah/blah/current. I guess sharp used bits and pieces from two different distros (one with the 'current' soft link and the other with the full directory name) and this confused it greatly. 
> 
> Solution: Make sure that old ccp4 distro is not sourced in any of your start-up shell scripts and no remnants of its path variable are left in your current PATH. Then reinstall sharp. 
> 
> Sincerely,
> 
> Petr
> 
> ------------------------
> Petr Leiman
> EPFL
> BSP-415
> CH-1015 Lausanne
> 
> ________________________________________
> From: sharp-discuss-bounces at globalphasing.com [sharp-discuss-bounces at globalphasing.com] on behalf of Luca Pellegrini [lp212 at cam.ac.uk]
> Sent: Tuesday, November 22, 2011 17:01
> To: sharp-discuss at globalphasing.com
> Subject: [sharp-discuss] Trivial problem?
> 
> Hello,
> 
> I am having an apparently trivial problem that I can't seem to figure out. I am trying to run Sharp in MAD mode, using peak (F_peak) and inflexion data (F_infl). When I start Sharp, it complains that it can't find Fs  of the correct type F in my file:
> 
> ERROR
>  : No column "F_peak" of type "F" found in file
>            /usr/local/sharp-2.6/users/lp212/None.sharp/datafiles/npi-esrf14.4.data.mtz!
> 
> As far as I can tell, the mtz file looks ok. I report below the mtzdump of my mtz file. Can anyone please help?
> 
> Thanks,
> Luca
> 
> ***mtzdump***
> 
>  OPENED INPUT MTZ FILE
>  Logical Name: HKLIN   Filename: /usr/local/sharp-2.6/users/lp212/None.sharp/datafiles/npi-esrf14.4.data.mtz
> 
> 
>  Spacegroup information obtained from library file:
>  Logical Name: SYMINFO   Filename: /sw/share/xtal/ccp4-6.2.0/lib/data/syminfo.lib
> 
>  * Title:
> 
>  npi esrf14.4 24october2011 Scala of ID29Oct11 H3 data
> 
>  * Base dataset:
> 
>         0 HKL_base
>           HKL_base
>           HKL_base
> 
>  * Number of Datasets = 3
> 
>  * Dataset ID, project/crystal/dataset names, cell dimensions, wavelength:
> 
>         1 New
>           New
>           New
>             154.7187  154.7187  145.2614   90.0000   90.0000  120.0000
>              0.99188
>         2 SeHM10N52
>           C5_9
>           Peak
>             152.5392  152.5392  141.8202   90.0000   90.0000  120.0000
>              0.97920
>         3 SeHM10N52
>           C5_9renamed
>           Infl
>             152.5300  152.5300  141.9500   90.0000   90.0000  120.0000
>              0.97950
> 
>  * Number of Columns = 17
> 
>  * Number of Reflections = 23536
> 
>  * Missing value set to NaN in input mtz file
> 
>  * HISTORY for current MTZ file :
> 
>  From FREERFLAG  6/10/2011 16:10:00 with fraction 0.050
>  data from CAD on  6/10/11
>  data from CAD on 18/11/11
> 
>  * Column Labels :
> 
>  H K L FreeR_flag F_nat SIGF_nat ISYM_nat F_peak SIGF_peak DANO_peak SIGDANO_peak ISYM_peak F_infl SIGF_infl DANO_infl SIGDANO_infl ISYM_infl
> 
>  * Column Types :
> 
>  H H H I F Q Y F Q D Q Y F Q D Q Y
> 
>  * Associated datasets :
> 
>  0 0 0 0 1 1 1 2 2 2 2 2 3 3 3 3 3
> 
>  * Cell Dimensions : (obsolete - refer to dataset cell dimensions above)
> 
>   154.7187  154.7187  145.2614   90.0000   90.0000  120.0000
> 
>  *  Resolution Range :
> 
>     0.00010    0.10905     (     98.489 -      3.028 A )
> 
>  * Sort Order :
> 
>       1     2     3     0     0
> 
>  * Space group = 'H 3' (number     146)
> 
>   (spacegroup is known)
> 
> 
> Luca Pellegrini
> Department of Biochemistry
> University of Cambridge
> 80 Tennis Court Road
> Cambridge CB2 1GA - UK
> 
> Email: lp212 at cam.ac.uk
> Tel: 0044-1223-760469
> Fax: 0044-1223-766002
> Sanger building, room 3.59
> 
> 
> 
> 
> 
> 
> 
> _______________________________________________
> sharp-discuss mailing list
> sharp-discuss at globalphasing.com
> https://www.globalphasing.com/mailman/listinfo/sharp-discuss
> _______________________________________________
> sharp-discuss mailing list
> sharp-discuss at globalphasing.com
> https://www.globalphasing.com/mailman/listinfo/sharp-discuss

-- 

***************************************************************
* Clemens Vonrhein, Ph.D.     vonrhein AT GlobalPhasing DOT com
*
*  Global Phasing Ltd.
*  Sheraton House, Castle Park 
*  Cambridge CB3 0AX, UK
*--------------------------------------------------------------
* BUSTER Development Group      (http://www.globalphasing.com)
***************************************************************


More information about the sharp-discuss mailing list