[gmx-users] Bug in g_confrms !

[Marc Baaden]

Hi,

we just found a bug in g_confrms. We have not yet spotted the origin of
the bug in the source code, but we can reproduce clearly wrong results.
The problems start from version 3.2.1, version 3.1.4 is ok. For those who
want to check, a simple test case is available here (just do make after
untarring):

http://www.shaman.ibpc.fr/rmsbug.tar

It seems like this happens when the number of atoms are different. The
example is just fitting 4 atoms on themselves, by randomly adding either
one atom, or one in two (eg fitting 5 on 4 atoms or 8 on 4). The option
-nomw amplifies the error. Below some numbers to show that the deviation
can be significant.

If anybody knows how to fix this, thanks in advance. Meanwhile, you should
probably revert back to gmx 314 for doing g_confrms type analysis.

Regards,
  Marc Baaden, Marie-Pierre Durrieu


::::::::::::::
gmx314.log      (no bug here)
::::::::::::::
The next four lines should show identical RMSDs
Root mean square deviation after lsq fit = 0.0265305
Root mean square deviation after lsq fit = 0.0265305
Root mean square deviation after lsq fit = 0.0265305
Root mean square deviation after lsq fit = 0.0265305
Let's prove they are identical to machine precision
Root mean square deviation after lsq fit = 3.02644e-08
Root mean square deviation after lsq fit = 3.02644e-08
Root mean square deviation after lsq fit = 3.02644e-08
Root mean square deviation after lsq fit = 3.02644e-08
Root mean square deviation after lsq fit = 3.02644e-08
Root mean square deviation after lsq fit = 3.02644e-08
::::::::::::::
gmx_mw.log      (slight deviations only)
::::::::::::::
The next four lines should show identical RMSDs
Root mean square deviation after lsq fit = 0.0265305
Root mean square deviation after lsq fit = 0.0280372
Root mean square deviation after lsq fit = 0.0280372
Root mean square deviation after lsq fit = 0.0286806
Let's prove they are identical to machine precision
Root mean square deviation after lsq fit = 3.32626e-05
Root mean square deviation after lsq fit = 3.32626e-05
Root mean square deviation after lsq fit = 3.63533e-05
Root mean square deviation after lsq fit = 2.1722e-08
Root mean square deviation after lsq fit = 5.1229e-05
Root mean square deviation after lsq fit = 5.1229e-05
::::::::::::::
gmx_nomw.log    (significant deviations with -nomw)
::::::::::::::
The next four lines should show identical RMSDs
Root mean square deviation after lsq fit = 0.0265305
Root mean square deviation after lsq fit = 0.0425423
Root mean square deviation after lsq fit = 0.366304
Root mean square deviation after lsq fit = 0.439791
Let's prove they are identical to machine precision
Root mean square deviation after lsq fit = 4.867e-05
Root mean square deviation after lsq fit = 3.63165e-05
Root mean square deviation after lsq fit = 3.63381e-05
Root mean square deviation after lsq fit = 4.20786e-05
Root mean square deviation after lsq fit = 4.21577e-05
Root mean square deviation after lsq fit = 5.0001e-05


-- 
 BioMolSim meeting 2&3 Sep 2005: http://www.iecb.u-bordeaux.fr/satellite2005/

 Dr. Marc Baaden  - Institut de Biologie Physico-Chimique, Paris
 mailto:baaden at smplinux.de      -      http://www.baaden.ibpc.fr
 FAX: +33 15841 5026  -  Tel: +33 15841 5176  ou  +33 609 843217