From: Ajasja Ljubetič (ajasja.ljubetic_at_gmail.com)
Date: Mon Feb 06 2012 - 15:04:56 CST
Dear Jrme and Giacomo,
> > Aha, I thought that it would take the indices of the refPositionsGroup
> and
> > only load those from the PDB.
>
> And you were right, it should, but it doesn't due to a bug that
> Giacomo just discovered (and that was my code too, mea culpa). That
> will work as documented in future revisions of the code.
>
> Great!
When a colvar depends on rotated coordinates, its gradient in the lab
> frame is the "counter-rotated" gradient in the rotated frame, plus
> another term that depends on derivatives of the optimal rotation with
> respect to Cartesian coordinates of the fitted atoms. That is handled
> explicitly in the RMSD and eigenvector coordinates, which are require
> fitted coordinates by nature, but it is awfully complicated to handle
> in a general way. So in other variables, the second term is neglected,
> and empirically it seems to work but it could become non-negligible in
> some non-intuitive cases.
>
> Thank you for the explanation. I don't think I've reached that problem
(yet). But I still have some other problems getting
this relative coordinate system to work:
To test rotateReference and centerReference I moved and rotated my helix as
shown here <http://lbf.ijs.si/ajasja/mails/ABF/relative-coords/S1.GIF>. I
left the reference dummy atoms untouched. The backbone atoms are still
fixed. I set the refPositionsFile to the "refrence" red helix (actually
just to the five atoms used for alignment <https://gist.github.com/1754767>).
This <http://lbf.ijs.si/ajasja/mails/ABF/relative-coords/S7.GIF> is the
PMF I obtained before and
this<http://lbf.ijs.si/ajasja/mails/ABF/relative-coords/S6.GIF> is
what I get using rotateReference and centerReference. So something is
still amiss.
To further track down the problem I calculated the values of theta and phi
(my colvars) form the trajectory using VMD (after aligning each frame to
the reference five heavy backbone atoms of the spin label). It turns out
the alignment is far from perfect (as shown
here<http://lbf.ijs.si/ajasja/mails/ABF/relative-coords/S2.GIF>),
But this should only "blur" the PMF .
It appears that the colvars that NAMD sees don't match what I get in VMD
(and what should be the correct values). Here is a
10ns<http://lbf.ijs.si/ajasja/mails/ABF/relative-coords/S4.GIF>
and 1ns <http://lbf.ijs.si/ajasja/mails/ABF/relative-coords/S5.GIF>
comparison. Is
the colvars module using the same algorithm for molecule alignment as VMD?
Because for the life of me, I can't figure out why the values don't match.
If I delete rotateRefrence and centerRefrence and use the refrence
(original) helix in the simulation everything is OK.
Thank you for your help & best regards,
Ajasja
PS: This <http://lbf.ijs.si/ajasja/mails/ABF/relative-coords/00001.colvars> is
the full colvar conf file.
This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:21:11 CST
