Hi Ugur, Mike had the same problem you had and solved it by issuing: matchHydrogenAtomLocations TRUE See the explanation below. Sorry it has taken so long to get back to you. I hope to have a real solution to this issue before too long, because reading hydrogen atom locations can cause trouble if the chirality is not clear (e.g. three bonds nearly coplanar, that sort of thing). but for now this may work fine for you, as it did for Mike. Sam
Samuel!
You are a champ! The option:
matchHydrogenAtomLocations True
seems to do the trick. I've been racking my brains on this one.
many thanks for the insight! Good luck with the grant writing! Never a fun time.
Best regards, Mike
On 03/04/13 21:01, Samuel Flores wrote:
Actually I think in this case the problem is more likely that molmodel has a bit of trouble creating atom positions for two hydrogens on a tetravalent atom, quite likely C5'. If the angle it forms with O5' and C4' is a bit obtuse, it can have trouble with this. Actually it's embarrassingly picky -- I've documented it puking when this angle is 153 degrees, which is really not far from the equilibrium angle. I am planning to make a more robust alternative method for this. Right now I am writing grants but have scheduled it for 22 April, with my apologies.
MMB by default ignores hydrogens, so you can try to get it to read hydrogen atom positions by setting
matchHydrogenAtomLocations True
.. though this very often causes additional problems, because recognizing chirality is itself an issue. So I'm not confident that this will solve your problem, though it might.
As a workaround you might try chopping out bits of your post-NAMD structure until MMB runs. This way you can identify which residue is problematic. moving the offending atom by a tenth of an Ångström might be enough to fix it. But I acknowledge that this is highly unattractive given the number of iterations you might end up going through.
Sam