Dear XPLOR-NIH mailinglist, I am Andras Czajlik from the Semmelweis University Budapest, and I am trying to use the Simulated Annealing Refinement module (refine.py) of XPLOR-NIH for protein structure refinement. Since I am a beginner in learning Xplor-NIH, I have several, quite basic questions about it:
1. The old user guide contains a sperate refinement (refine.inp) and analysis (accept.inp) section. Am I right that the refine.py includes both? 2. How does the XPLOR-NIH handle the distance restraints of those methyl (HG2* etc.), methylene (HB* etc) as well as aromatic protons (HE* etc) where there is only one observed chemical shift? Does it create a pseudoatom with averaged coordinates? If so, should I increase the upper bound of these restraints to make sure that the distances will be valid for the real atoms as well? 3. I noticed that the violation threshold for NOE restraints is 0.5A. Is it worth to try to do the refinement with lower (e.g. 0.2 or 0.3A) threshold value? If not, is there any option to print out all distances which violate the NOE upper distance bounds, even if the differences are between 0.0-0.5A? 4. I have done the structure calculation by Cyana, thus, I had to convert my data to XPLOR-NIH format. For this purpose, I used the PdbStat. I have read two guides about the conversion of the NOE distance restraints, which contradict with each other. One suggested to use the same upper bound as it was present in Cyana, the other advised to add further 10-15%. What do you think about it, which method would yield the better result? 5. During my first test I got a low number (2-4) hb violations. How important are they and how can I improve my structure? My experimental input data are the following: structure file (pdb), distance and backbone dihedral restraints. Many thanks in advance. Best regards, Andras ######################################################################## To unsubscribe from the XPLOR-NIH list, click the following link: http://list.nih.gov/cgi-bin/wa.exe?SUBED1=XPLOR-NIH&A=1
