[gmx-users] Bug in g_rms -> calculate RhoSc ??

[David van der Spoel]

On Tue, 2002-12-17 at 08:07, K.A. Feenstra wrote:
> Marc Baaden wrote:
> > Hi,
> > 
> > I was trying to use the alternative similarity measures Rho
> > and Rho Scaled from the Crippen and Maiorov paper. Rho looks
> > fine, but Rho scaled doesn't make sense, in the way that it
> > even looks like a bug to me ..
> 
> I haven't yet looked at your examples, but I was the one that implemented
> the rho/rho_sc measures in g_rms. I've definitely fixed some bugs before
> I was satisfied, and have checked several peptide conformations visually
> to see whether the calculated figures made any sense in comparison with
> the 'visual overlap' of the structures. Also, the behavior corresponds
> qualitatively to the description in the Crippen & Maiorov paper.
> 
> Other than that, there is no straightforward way of rigorously testing
> it, since I don't have reference structures to test against. Perhaps
> I (we?) could contact (one of) the authors to ask for a few structures
> with known rho/rho_sc differences. Do feel free, in the meantime, to
> check my implementation against the formulae in the paper. I did do
> some re-writing to get a more straightforward implementation.
Did you look at MArc's figures? It seemed like PBC errors to me... Or
maybe lsq fitting errors.


-- 
Groeten, David.
________________________________________________________________________
Dr. David van der Spoel, 	Biomedical center, Dept. of Biochemistry
Husargatan 3, Box 576,  	75123 Uppsala, Sweden
phone:	46 18 471 4205		fax: 46 18 511 755
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://xray.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++