Thomas Leeper wrote: > > 2) Regarding template generation, I've noticed that when I generate my > RNA template structure, about 1/2 to 2/3 of all of the amino H22s, 62s > or 42s, point back to the major groove face rather the the Watson-Crick > face. Is there some way to fix this so that they all point to the WC > face in the template? This would make using H-bond restraints derived > from HNN-COSY much more easy to implement. I'm currently using > topallhdg.dna and parallhdg.dna as topology and parameter files, seq2psf > to generate the psf file from sequence, and adding in the axis via > par_axis.pro and axis.psf during pdb construction. I've could try > editing the output pdb by hand, swithing H22 and H21 coordinates, but > that seems like a rather clumsy solution. I reckon that a DIHEdral > statement in the topology and parameter files could fix this > automatically but I'm not sure if I should edit these files, nor how > really. >
I just went back to the IUPAC definition (Eur. J. Biochem., vol 131, p. 9, fig. 2), and discovered that the improper torsion angle restraints for ade and cyt had the positions of H61|H62 (ade) and H41|H42 (cyt) reversed. I had changed these last March to conform to the output of another program (whose name is similar to a certain Space Shuttle). Those parameters are clearly in error. I've updated topallhdg.dna to reflect these corrections. Note that the definition in topallhdg.dna for the analagous situation in gua (H21|H22) was and is correct. Thanks for the catch. Producing correct topology/parameter sets is non-trivial, since it's so mind-numbingly easy to make a mistake. If other people have found mistakes in topallhdg.* or parallhdg.*, PLEASE let me know. It's the only way we'll all get correct parameters. --JK
