-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Hello Helene-- > I am running calculations on a modular protein (4 modules) of 270 amino > acids. I am trying to refine the NMR structure calculated by someoneelse > (with cyana followed with minimization with cns) using the same restraints > + 2 sets of dipolar couplings (NH only) in 2 different media (phage and > peg). I am using the refine.py script. So the tensors are allowed to vary, > but I have an initial guess from Module on the preliminary NMR structures. > > The sole changes I have made in the refine.py script concern the restraint > files. > The tensors were introduced with 700 and 701 numbering (as the residue > numbers for my protein begin at 355 and finish at 626. > This shouldn't matter as the psuedo atoms have a different segment name. > The modules are supposed to have only a limited conformational freedom, > and I have maintained the gyration potential (does -it make sense for an > elongated modular protein which is highly anisotropic? (estimated > anisotropy of 4)? If shape of the overall structure doesn't significantly differ from an ellipsoid and is expected to be well-packed then use of Vgyr should be safe. > *--- Dynamics ---- step= 0 ---- time= 0 ---- delta_t= > 0.001 --* > | E(kin)+E(poten)= nan E(kin)= 4341.138 temperature= > 2889.629 | > | E(poten)= nan grad= nan ANGL= > 66.7351079 | > | BOND= 1.7099773 CDIH= 99.7048545 HBDA= > nan | > | IMPR= 1.8225389 RAMA= -2.6377402 VDW= > 19.7891008 | > | Vgyr= 12721.0179159 noe= 33.4141355 rdc= > 129.1452028 | > *------------------------------------------------------------------------------* > "/PROG/LINUX/prog/Xplor-NIH/xplor-nih-2.26/python/wrappers/publicIVM.py", > line 136, in step > def step(self, *args, **kwargs): return > _publicIVM.PublicIVM_step(self, *args, **kwargs) > SystemError: xplor-nih error: eigen: 2 off-diagonal elements on an > intermediate tridiagonal form did not converge to zero. It seems that the IVM crashed due to a NAN in the HBDA term. You might place the following snippet before the first dyn.run() (line 346): from simulationTools import analyze print analyze(potList['HBDA']) Hopefully this will give some insight into why the HBDA term give a NAN. best regards-- Charles -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Processed by Mailcrypt 3.5.9 <http://mailcrypt.sourceforge.net/> iEYEARECAAYFAk0CP38ACgkQPK2zrJwS/lbDEACdEpYm4NMr810qpQDQxuB+Gael dH4An2wqQ/1A0KCwTP3A+whC6+u3UpvH =L4vR -----END PGP SIGNATURE-----
