Re: [ccp4bb] Bad geometry for alt. conformation refined in Refmac5

[Kay Diederichs]

Ian,

I cannot reproduce that bug on my x86_64 CentOS 5.4 system:

di...@turn29:-/tmp% cat temp.f90
      CHARACTER A(2)*2
      DATA A/2*'.A'/
      CALL CA(A)
      STOP 1
      END

      SUBROUTINE CA(A)
      CHARACTER A(2)*2
      DO 1 I=1,2
!#### With given data, the condition "A(I)(2:).NE.'A'" is always FALSE,
!#### so we should never execute the 'STOP 1' here:
1       IF(A(I)(2:).NE.'A' .AND. A(I).NE.'..') RETURN
!#### Should always drop through to here:
      STOP 2
      END

di...@turn29:-/tmp% ifort temp.f90
di...@turn29:-/tmp% ./a.out
2
di...@turn29:-/tmp% file a.out

di...@turn29:-/tmp% ifort -m32 temp.f90
di...@turn29:-/tmp% ./a.out
2
di...@turn29:-/tmp% file a.out

di...@turn29:-/tmp% ifort -v
Version 11.1
di...@turn29:-/tmp% ifort --vv
ifort: command line warning #10006: ignoring unknown option '-fvv'
ifort: command line error: no files specified; for help type "ifort -help"
di...@turn29:-/tmp% ifort -V

Copyright (C) 1985-2009 Intel Corporation.  All rights reserved.
FOR NON-COMMERCIAL USE ONLY

di...@turn29:-/tmp% uname -a


best,

Kay

Ian Tickle schrieb:

a.out: ELF 64-bit LSB executable, AMD x86-64, version 1 (SYSV), for GNU/Linux 2.6.9, dynamically linked (uses shared libs), for GNU/Linux 2.6.9, not stripped a.out: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), for GNU/Linux 2.6.9, dynamically linked (uses shared libs), for GNU/Linux 2.6.9, not stripped Intel(R) Fortran Intel(R) 64 Compiler Professional for applications running on Intel(R) 64, Version 11.1 Build 20091012 Package ID: l_cprof_p_11.1.059 Linux turn29.biologie.uni-konstanz.de 2.6.18-164.6.1.el5 #1 SMP Tue Nov 3 16:12:36 EST 2009 x86_64 x86_64 x86_64 GNU/Linux

Before you all go away and install the latest Intel compilers to fix
this bug - don't bother!  We just installed the latest version (ifort
11.3) and it has the same bug!  We plan to file a bug report with Intel.

In case anyone is interested, this is the code snippet which exhibits
the buggy behaviour (this was pruned down from s/r CHK_ALT in Refmac):

      CHARACTER A(2)*2
      DATA A/2*'.A'/
      CALL CA(A)
      STOP 1
      END

      SUBROUTINE CA(A)
      CHARACTER A(2)*2
      DO 1 I=2
C#### With given data, the condition "A(I)(2:).NE.'A'" is always FALSE,
C#### so we should never execute the 'STOP 1' here:
1       IF(A(I)(2:).NE.'A' .AND. A(I).NE.'..') RETURN
C#### Should always drop through to here:
      STOP 2
      END

It fails using -O2, -O3 and -Os, i.e. it only gives the right answer
using -O0 & -O1.

I should say that this is the first time we've ever encountered a bug in
the Intel compilers - I wish I could say the same for g77 & gfortran!

Cheers

-- Ian

-----Original Message-----

From: owner-ccp...@jiscmail.ac.uk [mailto:owner-ccp...@jiscmail.ac.uk] On Behalf Of Ian Tickle

Sent: 28 November 2009 13:22
To: Garib Murshudov; john.pas...@jefferson.edu
Cc: CCP4BB@jiscmail.ac.uk


I think I've finally squashed this particular bug which had been
annoying me for quite some time!  It appears that Intel Fortran v11.0
has an optimisation bug in s/r CHK_ALT (i.e. checking alternate atom

& possibly earlier seem to be bug-free (at least this bug anyway!).
Possibly this is fixed in the current version of ifort (11.1), again I

are unaffected.

So my solution is to recompile with -O1.  The CCP4 6.1.2 distribution
uses -O2 so you need either to fix the makefile, or download Garib's

settings.


an issue, because at least for the OS I'm using (Centos 4.6),
gfortran-compiled code runs ~ 300% slower than the ifort code!!

Hope this helps!

-- Ian

Subject: RE: [ccp4bb] Bad geometry for alt. conformation refined in Refmac5 codes), & possibly also 10.x - I don't have those to test - though v9.1 don't have it to test, maybe someone who does can try it. This affects compilations using optimisation levels -O2, -Os and -O3, but -O0 and -O1 source code and make sure the makefile you use has the appropriate flag As was pointed out earlier, having to use the gfortran-compiled code is

-----Original Message-----

From: owner-ccp...@jiscmail.ac.uk [mailto:owner-ccp...@jiscmail.ac.uk] On Behalf Of Garib Murshudov

Sent: 25 November 2009 16:27
To: john.pas...@jefferson.edu
Cc: CCP4BB@jiscmail.ac.uk


Could you please try the version from York:

www.ysbl.york.ac.uk/refmac/latest_refmac.html



regards
Garib

On 25 Nov 2009, at 16:18, John Pascal wrote:

Subject: Re: [ccp4bb] Bad geometry for alt. conformation refined in Refmac5 I think probkem you mention is related with compilation or something. At least I cannot repat it on my computer

Hello All,


geometry restraints are not well defined.

We are trying to refine ARG residues with two conformations in Refmac5, and the refined atom positions in the output PDB file are all over the place, as if the We've tried several different formats for the input file,

based on

previous

postings to the bulletin board and the PDB standard (two

examples

below), but
the result is always the same.

We are using Refmac5 in CCP4 Suite 6.1.2, GUI 2.0.5 on Mac OSX.

We'd appreciate any suggestions.  Thanks.  -John

Examples of ARG format:

1)


2)




John Pascal, PhD
Thomas Jefferson University
Department of Biochemistry & Molecular Biology
233 South 10th Street, BLSB 804
Philadelphia, Pennsylvania  19107

ph 215.503.4596
fx  215.923.2117

ATOM 1472 N ARG A 1 -5.737 26.887 38.372 1.00 29.53 C N ATOM 1473 CA ARG A 1 -5.445 25.560 37.882 1.00 30.24 C C ATOM 1474 CB ARG A 1 -5.314 24.548 39.036 1.00 30.63 C C ATOM 1475 CG ARG A 1 -5.426 23.052 38.627 1.00 34.81 C C ATOM 1476 CD AARG A 1 -4.827 22.075 39.644 0.50 37.09 C C ATOM 1477 CD BARG A 1 -4.301 22.419 39.279 0.50 37.09 C C ATOM 1478 NE AARG A 1 -3.430 21.777 39.304 0.50 42.71 C N ATOM 1479 NE BARG A 1 -4.482 21.902 40.627 0.50 42.71 C N ATOM 1480 CZ AARG A 1 -2.998 20.868 38.402 0.50 44.91 C C ATOM 1481 CZ BARG A 1 -3.648 22.142 41.638 0.50 44.91 C C ATOM 1482 NH1AARG A 1 -3.841 20.117 37.678 0.50 45.20 C N ATOM 1483 NH1BARG A 1 -2.584 22.912 41.464 0.50 45.20 C N ATOM 1484 NH2AARG A 1 -1.688 20.715 38.210 0.50 44.99 C N ATOM 1485 NH2BARG A 1 -3.878 21.619 42.831 0.50 44.99 C N ATOM 1486 C ARG A 1 -6.518 25.176 36.830 1.00 29.97 C C ATOM 1487 O ARG A 1 -7.675 25.501 36.971 1.00 31.02 C O ATOM 44 N AARG A 1 26.671 62.112 46.990 0.50 30.13 A N ATOM 45 CA AARG A 1 26.970 63.346 47.667 0.50 30.65 A C ATOM 46 CB AARG A 1 27.172 64.495 46.676 0.50 31.07 A C ATOM 47 CG AARG A 1 27.152 65.897 47.322 0.50 34.20 A C ATOM 48 CD AARG A 1 27.993 66.976 46.599 0.50 37.16 A C ATOM 49 NE AARG A 1 27.726 67.425 45.342 0.50 42.06 A N ATOM 50 CZ AARG A 1 28.315 67.639 44.168 0.50 44.78 A C ATOM 51 NH1AARG A 1 29.525 67.160 43.918 0.50 45.37 A N ATOM 52 NH2AARG A 1 27.690 68.340 43.240 0.50 45.67 A N ATOM 53 C AARG A 1 25.839 63.640 48.622 0.50 30.37 A C ATOM 54 O AARG A 1 24.690 63.377 48.340 0.50 31.24 A O ATOM 55 N BARG A 1 26.667 62.080 47.010 0.50 30.13 A N ATOM 56 CA BARG A 1 26.921 63.329 47.640 0.50 30.65 A C ATOM 57 CB BARG A 1 27.108 64.390 46.581 0.50 31.07 A C ATOM 58 CG BARG A 1 27.138 65.756 47.103 0.50 34.20 A C ATOM 59 CD BARG A 1 28.447 66.452 46.933 0.50 37.16 A C ATOM 60 NE BARG A 1 28.377 67.657 47.707 0.50 42.06 A N ATOM 61 CZ BARG A 1 29.349 68.269 48.373 0.50 44.78 A C ATOM 62 NH1BARG A 1 30.572 67.836 48.373 0.50 45.37 A N ATOM 63 NH2BARG A 1 29.066 69.368 49.034 0.50 45.67 A N ATOM 64 C BARG A 1 25.822 63.636 48.638 0.50 30.37 A C ATOM 65 O BARG A 1 24.698 63.372 48.403 0.50 31.24 A O

Disclaimer

This communication is confidential and may contain privileged information intended solely for the named addressee(s). It may not be used or disclosed except for the purpose for which it has been sent. If you are not the intended recipient you must not review, use, disclose, copy, distribute or take any action in reliance upon it. If you have received this communication in error, please notify Astex Therapeutics Ltd by emailing i.tic...@astex-therapeutics.com and destroy all copies of the message and any attached documents. Astex Therapeutics Ltd monitors, controls and protects all its messaging traffic in compliance with its corporate email policy. The Company accepts no liability or responsibility for any onward transmission or use of emails and attachments having left the Astex Therapeutics domain. Unless expressly stated, opinions in this message are those of the individual sender and not of Astex Therapeutics Ltd. The recipient should check this email and any attachments for the presence of computer viruses. Astex Therapeutics Ltd accepts no liability for damage caused by any virus transmitted by this email. E-mail is susceptible to data corruption, interception, unauthorized amendment, and tampering, Astex Therapeutics Ltd only send and receive e-mails on the basis that the Company is not liable for any such alteration or any consequences thereof. Astex Therapeutics Ltd., Registered in England at 436 Cambridge Science Park, Cambridge CB4 0QA under number 3751674

Disclaimer

Astex Therapeutics Ltd., Registered in England at 436 Cambridge Science Park, 
Cambridge CB4 0QA under number 3751674

This communication is confidential and may contain privileged information intended solely for the named addressee(s). It may not be used or disclosed except for the purpose for which it has been sent. If you are not the intended recipient you must not review, use, disclose, copy, distribute or take any action in reliance upon it. If you have received this communication in error, please notify Astex Therapeutics Ltd by emailing i.tic...@astex-therapeutics.com and destroy all copies of the message and any attached documents. Astex Therapeutics Ltd monitors, controls and protects all its messaging traffic in compliance with its corporate email policy. The Company accepts no liability or responsibility for any onward transmission or use of emails and attachments having left the Astex Therapeutics domain. Unless expressly stated, opinions in this message are those of the individual sender and not of Astex Therapeutics Ltd. The recipient should check this email and any attachments for the presence of computer viruses. Astex Therapeutics Ltd accepts no liability for damage caused by any virus transmitted by this email. E-mail is susceptible to data corruption, interception, unauthorized amendment, and tampering, Astex Therapeutics Ltd only send and receive e-mails on the basis that the Company is not liable for any such alteration or any consequences thereof.

--
Kay Diederichs                http://strucbio.biologie.uni-konstanz.de
email: kay.diederi...@uni-konstanz.de    Tel +49 7531 88 4049 Fax 3183
Fachbereich Biologie, Universität Konstanz, Box M647, D-78457 Konstanz

This e-mail is digitally signed. If your e-mail client does not have the
necessary capabilities, just ignore the attached signature "smime.p7s".

smime.p7s
Description: S/MIME Cryptographic Signature